Recurrent neural network grammars. Slide credits: Chris Dyer, Adhiguna Kuncoro
|
|
- Jason McDonald
- 5 years ago
- Views:
Transcription
1 Recurrent neural network grammars Slide credits: Chris Dyer, Adhiguna Kuncoro
2 Widespread phenomenon: Polarity items can only appear in certain contexts Example: anybody is a polarity item that tends to appear only in specific contexts: The lecture that I gave did not appeal to anybody but not: * The lecture that I gave appealed to anybody We might infer that the licensing context is the word not appearing among the preceding words, and you could use an RNN to model this. However: * The lecture that I did not give appealed to anybody
3 Language is hierarchical The licensing context depends on recursive structure (syntax) The lecture did not that I gave appeal to anybody The lecture appealed to anybody that I did not give
4 One theory of hierarchy Generate symbols sequentially using an RNN Add some control symbols to rewrite the history periodically Periodically compress a sequence into a single constituent Augment RNN with an operation to compress recent history into a single vector (-> reduce ) RNN predicts next symbol based on the history of compressed elements and non-compressed terminals ( push ) RNN must also predict control symbols that decide how big constituents are We call such models recurrent neural network grammars.
5 One theory of hierarchy Generate symbols sequentially using an RNN Add some control symbols to rewrite the history periodically Periodically compress a sequence into a single constituent Augment RNN with an operation to compress recent history into a single vector (-> reduce ) RNN predicts next symbol based on the history of compressed elements and non-compressed terminals ( shift or generate ) RNN must also predict control symbols that decide how big constituents are We call such models recurrent neural network grammars.
6 One theory of hierarchy Generate symbols sequentially using an RNN Add some control symbols to rewrite the history periodically Periodically compress a sequence into a single constituent Augment RNN with an operation to compress recent history into a single vector (-> reduce ) RNN predicts next symbol based on the history of compressed elements and non-compressed terminals ( shift or generate ) RNN must also predict control symbols that decide how big constituents are We call such models recurrent neural network grammars.
7 (Ordered) tree traversals are sequences S NP VP The hungry cat meows.
8 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
9 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
10 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
11 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
12 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
13 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
14 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
15 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
16 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
17 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
18 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
19 (Ordered) tree traversals are sequences S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). )
20 Terminals Stack Action
21 Terminals Stack Action NT(S)
22 Terminals Stack Action NT(S) (S NT(NP)
23 Terminals Stack Action NT(S) (S (NP (S NT(NP)
24 Terminals Stack Action NT(S) (S NT(NP) (S (NP GEN(The)
25 Terminals Stack Action NT(S) The (S NT(NP) (S (NP GEN(The) (S (NP The
26 Terminals Stack Action NT(S) The (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry)
27 Terminals Stack Action NT(S) The The hungry (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry
28 Terminals Stack Action NT(S) The The hungry (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat)
29 Terminals Stack Action NT(S) The The hungry The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat
30 Terminals Stack Action NT(S) The The hungry The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE
31 Terminals Stack Action NT(S) (S NT(NP) (S (NP GEN(The) The (S (NP The GEN(hungry) The hungry (S (NP The hungry GEN(cat) The hungry cat (S (NP The hungry cat REDUCE The hungry cat (S (NP The hungry cat )
32 Terminals Stack Action NT(S) (S NT(NP) (S (NP GEN(The) The The hungry The hungry cat (S (NP The (S (NP The hungry (S (NP The hungry cat GEN(hungry) GEN(cat) REDUCE The hungry cat (S (NP The hungry cat ) (S (NP The hungry cat) Compress The hungry cat into a single composite symbol
33 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE (S (NP The hungry cat)
34 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE (S (NP The hungry cat) NT(VP)
35 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE (S (NP The hungry cat) NT(VP) (S (NP The hungry cat) (VP
36 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE (S (NP The hungry cat) NT(VP) The hungry cat (S (NP The hungry cat) (VP??? Q: What information can we use to predict the next action, and how can we encode it with an RNN?
37 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE (S (NP The hungry cat) NT(VP) (S (NP The hungry cat) (VP A: We can use an RNN for each of: 1. Previous terminal symbols 2. Previous actions 3. Current stack contents
38 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE (S (NP The hungry cat) NT(VP) The hungry cat (S (NP The hungry cat) (VP GEN(meows)
39 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat (S NT(NP) (S (NP GEN(The) (S (NP The GEN(hungry) (S (NP The hungry GEN(cat) (S (NP The hungry cat REDUCE (S (NP The hungry cat) NT(VP) The hungry cat The hungry cat meows (S (NP The hungry cat) (VP (S (NP The hungry cat) (VP meows GEN(meows) REDUCE The hungry cat meows (S (NP The hungry cat) (VP meows) GEN(.) The hungry cat meows. The hungry cat meows. (S (NP The hungry cat) (VP meows). (S (NP The hungry cat) (VP meows).) REDUCE
40 Terminals Stack Action NT(S) The The hungry The hungry cat The hungry cat The hungry cat The hungry cat meows The hungry cat meows (S (NP The (S (NP The hungry (S (NP The hungry cat (S (NP The hungry cat) Final stack symbol is (a vector representation of) the complete tree. (S (NP The hungry cat) (VP (S (NP The hungry cat) (VP meows (S NT(NP) (S (NP GEN(The) GEN(hungry) GEN(cat) REDUCE NT(VP) GEN(meows) REDUCE (S (NP The hungry cat) (VP meows) GEN(.) The hungry cat meows. The hungry cat meows. (S (NP The hungry cat) (VP meows). (S (NP The hungry cat) (VP meows).) REDUCE
41 Syntactic Composition Need representation for: (NP The hungry cat)
42 Syntactic Composition Need representation for: (NP The hungry cat) What head type? NP The
43 Syntactic Composition Need representation for: (NP The hungry cat) What head type? NP The hungry
44 Syntactic Composition Need representation for: (NP The hungry cat) What head type? NP The hungry cat
45 Syntactic Composition Need representation for: (NP The hungry cat) What head type? NP The hungry cat )
46 Syntactic Composition Need representation for: (NP The hungry cat) NP The hungry cat )
47 Syntactic Composition Need representation for: (NP The hungry cat) NP The hungry cat ) NP
48 Syntactic Composition Need representation for: (NP The hungry cat) NP The hungry cat ) NP
49 Syntactic Composition Need representation for: (NP The hungry cat) NP The hungry cat ) NP
50 Syntactic Composition Need representation for: (NP The hungry cat) NP The hungry cat ) NP
51 Syntactic Composition Need representation for: (NP The hungry cat) ( NP The hungry cat ) NP
52 Recursion Need representation for: (NP The hungry cat) (NP The (ADJP very hungry) cat) ( The hungry NP cat ) NP
53 Recursion Need representation for: (NP The hungry cat) (NP The (ADJP very hungry) cat) {z } v ( v NP The cat ) NP
54 Recursion Need representation for: (NP The hungry cat) (NP The (ADJP very hungry) cat) {z } v ( v NP The cat ) NP
55 Stack symbols composed recursively mirror corresponding tree structure S NP VP The hungry cat meows. The hungry cat meows.
56 Stack symbols composed recursively mirror corresponding tree structure S NP VP The hungry cat meows. The hungry cat meows. NP
57 Stack symbols composed recursively mirror corresponding tree structure S NP VP The hungry cat meows. The hungry cat meows. NP VP
58 Stack symbols composed recursively mirror corresponding tree structure S Effect Stack encodes NP VP top-down syntactic recency, rather The hungry cat meows. than left-to-right string recency The hungry cat meows. NP VP S
59 Implementing RNNGs Stack RNNs Augment a sequential RNN with a stack pointer Two constant-time operations push - read input, add to top of stack, connect to current location of the stack pointer pop - move stack pointer to its parent A summary of stack contents is obtained by accessing the output of the RNN at location of the stack pointer Note: push and pop are discrete actions here (cf. Grefenstette et al., 2015)
60 Implementing RNNGs Stack RNNs y 0 PUSH ;
61 Implementing RNNGs Stack RNNs y 0 y 1 POP ; x 1
62 Implementing RNNGs Stack RNNs y 0 y 1 ; x 1
63 Implementing RNNGs Stack RNNs y 0 y 1 PUSH ; x 1
64 Implementing RNNGs Stack RNNs y 0 y 1 y 2 POP ; x 1 x 2
65 Implementing RNNGs Stack RNNs y 0 y 1 y 2 ; x 1 x 2
66 Implementing RNNGs Stack RNNs y 0 y 1 y 2 PUSH ; x 1 x 2
67 Implementing RNNGs Stack RNNs y 0 y 1 y 2 y 3 ; x 1 x 2 x 3
68 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) stack top
69 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) stack S top
70 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) NP top stack S
71 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) NP top stack S
72 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) top NP stack S
73 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) top NP stack S
74 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) NP top stack S
75 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) NP VP top stack S
76 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) top NP VP stack S
77 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) NP VP top stack S
78 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) top NP VP stack S
79 The evolution of the stack LSTM over time mirrors tree structure S NP VP The hungry cat meows. S( NP( The hungry cat ) VP( meows ). ) NP VP stack S top
80 Each word is conditioned on history represented by a trio of RNNs S NP VP The hungry cat meows. p(meows history) S( NP( The hungry cat ) VP( meows ). ) NP VP stack S
81 Train with backpropagation through In training, backpropagate through these three RNNs) structure S NP VP The hungry cat meows. This network is dynamic. Don t derive gradients by hand that s error prone. Use automatic differentiation instead And recursively through this structure. stack S( NP( The hungry cat ) VP( meows ). ) NP VP S
82 Complete model sentence Sequence of actions (completely defines x and y) Actions up to time t tree bias allowable actions at this step action embedding history embedding
83 Complete model sentence Sequence of actions (completely defines x and y) Actions up to time t tree bias Model is dynamic: variable number of context-dependent actions at each step allowable actions at this step action embedding history embedding
84 Complete model stack output (buffer) action history
85 Complete model stack output (buffer) action history
86 Implementing RNNGs Parameter Estimation RNNGs jointly model sequences of words together with a tree structure, p (x, y) Any parse tree can be converted to a sequence of actions (depth first traversal) and vice versa (subject to wellformedness constraints) We use trees from the Penn Treebank We could treat the non-generation actions as latent variables or learn them with RL, effectively making this a problem of grammar induction. Future work
87 Implementing RNNGs Inference An RNNG is a joint distribution p(x,y) over strings (x) and parse trees (y) We are interested in two inference questions: What is p(x) for a given x? [language modeling] What is max p(y x) for a given x? [parsing] y Unfortunately, the dynamic programming algorithms we often rely on are of no help here We can use importance sampling to do both by sampling from a discriminatively trained model
88 English PTB (Parsing) Type F1 Petrov and Klein (2007) Shindo et al (2012) Single model Shindo et al (2012) Ensemble Vinyals et al (2015) PTB only Vinyals et al (2015) Ensemble G 90.1 G 91.1 ~G 92.4 D 90.5 S 92.8 Discriminative D 89.8 Generative (IS) G 92.4
89 Importance Sampling Assume we ve got a conditional distribution q(y x) s.t. (i) p(x, y) > 0 =) q(y x) > 0 (ii) y q(y x) is tractable and (iii) q(y x) is tractable
90 Importance Sampling Assume we ve got a conditional distribution q(y x) s.t. (i) p(x, y) > 0 =) q(y x) > 0 (ii) y q(y x) is tractable and (iii) q(y x) is tractable Let the importance weights w(x, y) = p(x, y) q(y x)
91 Importance Sampling Assume we ve got a conditional distribution q(y x) s.t. (i) p(x, y) > 0 =) q(y x) > 0 (ii) y q(y x) is tractable and (iii) q(y x) is tractable Let the importance weights w(x, y) = p(x, y) q(y x) p(x) = X p(x, y) = X w(x, y)q(y x) y2y(x) y2y(x) = E y q(y x) w(x, y)
92 Importance Sampling p(x) = X p(x, y) = X w(x, y)q(y x) y2y(x) y2y(x) = E y q(y x) w(x, y)
93 Importance Sampling p(x) = X p(x, y) = X w(x, y)q(y x) y2y(x) y2y(x) = E y q(y x) w(x, y) Replace this expectation with its Monte Carlo estimate. y (i) q(y x) for i 2 {1, 2,...,N}
94 Importance Sampling p(x) = X p(x, y) = X w(x, y)q(y x) y2y(x) y2y(x) = E y q(y x) w(x, y) Replace this expectation with its Monte Carlo estimate. y (i) q(y x) for i 2 {1, 2,...,N} E q(y x) w(x, y) MC 1 N NX w(x, y (i) ) i=1
95 English PTB (LM) Perplexity 5-gram IKN LSTM + Dropout Generative (IS) Chinese CTB (LM) Perplexity 5-gram IKN LSTM + Dropout Generative (IS) 171.9
96 Do we need a stack? Kuncoro et al., Oct 2017 Both stack and action history encode the same information, but expose it to the classifier in different ways. Leaving out stack is harmful; using it on its own works slightly better than complete model!
97 RNNG as a mini-linguist Replace composition with one that computes attention over objects in the composed sequence, using embedding of NT for similarity. What does this learn?
98 RNNG as a mini-linguist Replace composition with one that computes attention over objects in the composed sequence, using embedding of NT for similarity. What does this learn?
99 RNNG as a mini-linguist Replace composition with one that computes attention over objects in the composed sequence, using embedding of NT for similarity. What does this learn?
100 RNNG as a mini-linguist Replace composition with one that computes attention over objects in the composed sequence, using embedding of NT for similarity. What does this learn?
101 RNNG as a mini-linguist Replace composition with one that computes attention over objects in the composed sequence, using embedding of NT for similarity. What does this learn?
102 Summary Language is hierarchical, and this inductive bias can be encoded into an RNN-style model. RNNGs work by simulating a tree traversal like a pushdown automaton, but with continuous rather than finite history. Modeled by RNNs encoding (1) previous tokens, (2) previous actions, and (3) stack contents. A stack LSTM evolves with stack contents. The final representation computed by a stack LSTM has a topdown recency bias, rather than left-to-right bias, which might be useful in modeling sentences. Effective for parsing and language modeling, and seems to capture linguistic intuitions about headedness.
IMAGE CAPTIONING USING PHRASE-BASED HIERARCHICAL LSTM MODEL
IMAGE CAPTIONING USING PHRASE-BASED HIERARCHICAL LSTM MODEL 1 Chee Seng Chan PhD SMIEEE 23 October 2017 Nvidia AI Conference, Singapore email: cs.chan@um.edu.my INTRODUCTION Aim: Automatic generate a full
More informationDynamic Programming for Linear Time Incremental Parsing
Dynamic Programming for Linear Time ncremental Parsing Liang Huang nformation Sciences nstitute University of Southern California Kenji Sagae nstitute for Creative Technologies University of Southern California
More informationThe online processing of semantic and pragmatic content
The online processing of semantic and pragmatic content LINGUIST510 Brian Dillon Psycholinguistics Comprehension: How do we compute the meaning of a sentence in real time? What are the online computations
More informationGenera&on of Image Descrip&ons. Tambet Ma&isen
Genera&on of Image Descrip&ons Tambet Ma&isen 14.10.2015 Agenda Datasets Convolu&onal neural networks Neural language models Neural machine transla&on Genera&on of image descrip&ons AFen&on Metrics A
More informationMoving toward formalisation COMP62342
Moving toward formalisation COMP62342 Sean Bechhofer sean.bechhofer@manchester.ac.uk Uli Sattler uli.sattler@manchester.ac.uk (thanks to Bijan Parsia for slides) Previously... We started the knowledge
More informationMachine Learning.! A completely different way to have an. agent acquire the appropriate abilities to solve a particular goal is via machine learning.
Machine Learning! A completely different way to have an agent acquire the appropriate abilities to solve a particular goal is via machine learning. Machine Learning! What is Machine Learning? " Programs
More informationReasoning with Neural Networks
University of Toronto Faculty of Arts and Science Department of Computer Science Reasoning with Neural Networks Rodrigo Toro Icarte (rntoro@cs.toronto.edu) March 08, 2016 Motivation Could a crocodile run
More informationOn Deriving Aspectual Sense
COGNrTlVE SCIENCE 2,385-390 (1978) On Deriving Aspectual Sense BONNIE LYNN WEBBER University of Pennsylvania and Bolt Beranek and Newrnan. Inc. In his recent article "Verbs, Time and Modality," M. J. Steedman
More informationDevelopment of intelligent systems (RInS) Object recognition with Convolutional Neural Networks
Development of intelligent systems (RInS) Object recognition with Convolutional Neural Networks Danijel Skočaj University of Ljubljana Faculty of Computer and Information Science Academic year: 2017/18
More informationNatural Language Processing (NLP)
Natural Language Processing (NLP) Goal: Understand the meaning of natural language Applications Information retrieval Machine translation Dialogue systems Example: IBM Watson in quiz show NLP is difficult
More informationSemantics. These slides were produced by Hadas Kotek.
Semantics These slides were produced by Hadas Kotek. http://web.mit.edu/hkotek/www/ 1 Sentence types What is the meaning of a sentence? The lion devoured the pizza. Statement 2 Sentence types What is the
More informationWeek 42: Siamese Network: Architecture and Applications in Visual Object Tracking. Yuanwei Wu
Week 42: Siamese Network: Architecture and Applications in Visual Object Tracking Yuanwei Wu 10-21-2016 1 Outline Siamese Architecture Siamese Applications in Computer Vision Paper review Visual Object
More informationWhat kind of Theory do we need for English Syntax? Are languages finite? Could we list all the sentences of English?
What kind of Theory do we need for English Syntax? Are languages finite? Could we list all the sentences of English? Some sentences go on and on. Some sentences go on and on and on. Some sentences go on
More informationBayesian Analysis of Population Mixture and Admixture
Bayesian Analysis of Population Mixture and Admixture Eric C. Anderson Interdisciplinary Program in Quantitative Ecology and Resource Management University of Washington, Seattle, WA, USA Jonathan K. Pritchard
More informationAn Introduction to Formal Logic
An Introduction to Formal Logic Richard L. Epstein Advanced Reasoning Forum Copyright 2016 by Richard L. Epstein. All rights reserved. No part of this work may be reproduced, stored in a retrieval system,
More informationCS6501: Deep Learning for Visual Recognition. CNN Architectures
CS6501: Deep Learning for Visual Recognition CNN Architectures ILSVRC: ImagenetLarge Scale Visual Recognition Challenge [Russakovsky et al 2014] The Problem: Classification Classify an image into 1000
More informationLogical Forms. Prof. Sameer Singh CS 295: STATISTICAL NLP WINTER February 16, 2017
Logical Forms Prof. Sameer Singh CS 295: STATISTICAL NLP WINTER 2017 February 16, 2017 Based on slides from Noah Smith, Dan Klein, Tom Kwiatkowski, and everyone else they copied from. Outline Logical Semantics
More informationNathan A. Thompson, Ph.D. Adjunct Faculty, University of Cincinnati Vice President, Assessment Systems Corporation
An Introduction to Computerized Adaptive Testing Nathan A. Thompson, Ph.D. Adjunct Faculty, University of Cincinnati Vice President, Assessment Systems Corporation Welcome! CAT: tests that adapt to each
More informationMoving towards formalisation COMP62342
Moving towards formalisation COMP62342 Sean Bechhofer sean.bechhofer@manchester.ac.uk Uli Sattler uli.sattler@manchester.ac.uk (thanks to Bijan Parsia for slides) Previously... We started the Knowledge
More informationDesign of 32 bit Parallel Prefix Adders
IOSR Journal of Electronics and Communication Engineering (IOSR-JECE) e-issn: 2278-2834,p- ISSN: 2278-8735. Volume 6, Issue 1 (May. - Jun. 2013), PP 01-06 Design of 32 bit Parallel Prefix Adders P.Chaitanya
More informationMulticlass and Multi-label Classification
Multiclass and Multi-label Classification INFO-4604, Applied Machine Learning University of Colorado Boulder September 21, 2017 Prof. Michael Paul Today Beyond binary classification All classifiers we
More informationCourse # Course Name Credits
Curriculum Outline: Course # Course Name Credits Term 1 Courses VET 100 Introduction to Veterinary Technology 3 ENG 105 English Composition 3 MATH 120 Technical Mathematics 3 VET 130 Animal Biology/ Anatomy
More informationLecture 1: Turtle Graphics. the turtle and the crane and the swallow observe the time of their coming; Jeremiah 8:7
Lecture 1: Turtle Graphics the turtle and the crane and the sallo observe the time of their coming; Jeremiah 8:7 1. Turtle Graphics The turtle is a handy paradigm for the study of geometry. Imagine a turtle
More informationUsing Physics for Motion Retargeting
Thesis Submitted to Utrecht University for the degree of Master of Science Supervisor: drs. Arno Kamphuis INF/SCR-10-13 Utrecht University Department of Computer Science MSc Program: Game and Media Technology
More informationFunctions Introduction to Functions 7.2 One-to-One, Onto, Inverse functions. mjarrar Watch this lecture and download the slides
9/6/17 Mustafa Jarrar: Lecture Notes in Discrete Mathematics Birzeit University Palestine 2015 Functions 71 Introduction to Functions 72 One-to-One Onto Inverse functions 73 Application: The Pigeonhole
More informationContents. Resources. Download these resources in your computer/laptop from
Contents Resources... 2 Activities... 3 Worksheets (Beginner and Intermediate) Auslan Auslan Auslan Auslan Auslan Bird Story cut and paste... Deer Story cut and paste... Caterpillar Story cut and paste...
More informationMr. Bouchard Summer Assignment AP Biology. Name: Block: Score: / 20. Topic: Chemistry Review and Evolution Intro Packet Due: 9/4/18
Name: Block: Score: / 20 Topic: Chemistry Review and Evolution Intro Packet Due: 9/4/18 Week Schedule Monday Tuesday Wednesday Thursday Friday In class discussion/activity NONE NONE NONE Syllabus and Course
More information[Boston March for Science 2017 photo Hendrik Strobelt]
[Boston March for Science 2017 photo Hendrik Strobelt] [Boston March for Science 2017] [Boston March for Science 2017] [Boston March for Science 2017] Object Detectors Emerge in Deep Scene CNNs Bolei
More informationEating Your Own Dog Food
Eating Your Own Dog Food Andreas Abel 1 1 Department of Computer Science and Engineering Chalmers and Gothenburg University, Sweden 25th Agda Implementor s Meeting (AIM XXI) Teknikparken, Chalmers, Gothenburg,
More informationControllability of Complex Networks. Yang-Yu Liu, Jean-Jacques Slotine, Albert-Laszlo Barbasi Presented By Arindam Bhattacharya
Controllability of Complex Networks Yang-Yu Liu, Jean-Jacques Slotine, Albert-Laszlo Barbasi Presented By Arindam Bhattacharya Index Overview Network Controllability Controllability of real networks An
More informationApplication of Fuzzy Logic in Automated Cow Status Monitoring
University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln Biological Systems Engineering: Papers and Publications Biological Systems Engineering 2001 Application of Fuzzy Logic in
More informationPhysics Based Ragdoll Animation
Physics Based Ragdoll Animation Arash Ghodsi & David Wilson Abstract Ragdoll animation is a technique used to add realism to falling bodies with multiple joints, such as a human model. Doing it right can
More informationNATURAL SELECTION SIMULATION
ANTHR 1-L BioAnthro Lab Name: NATURAL SELECTION SIMULATION INTRODUCTION Natural selection is an important process underlying the theory of evolution as proposed by Charles Darwin and Alfred Russell Wallace.
More informationProf Michael O Neill Introduction to Evolutionary Computation
Prof Michael O Neill Introduction to Evolutionary Computation Origin of the Species Million Years Ago Event? Origin of Life 3500 Bacteria 1500 Eukaryotic Cells 600 Multicellular Organisms 1 Human Language
More informationReversing Category Exclusivities in Infant Perceptual Categorization: Simulations and Data
Reversing Category Exclusivities in Infant Perceptual Categorization: Simulations and Data Robert M. French, Martial Mermillod (rfrench, mmermillod@ulg.ac.be) Psychology Department, Université de Liège,
More informationSuperlative Quantifiers as Meta Speech Acts
Superlative Quantifiers as Meta Speech Acts Ariel Cohen Manfred Krifka Ben Gurion University of the Negev Humboldt-Universität zu Berlin & Zentrum für Allgemeine Sprachwissenschaft, Berlin Introduction
More informationSEDAR31-DW30: Shrimp Fishery Bycatch Estimates for Gulf of Mexico Red Snapper, Brian Linton SEDAR-PW6-RD17. 1 May 2014
SEDAR31-DW30: Shrimp Fishery Bycatch Estimates for Gulf of Mexico Red Snapper, 1972-2011 Brian Linton SEDAR-PW6-RD17 1 May 2014 Shrimp Fishery Bycatch Estimates for Gulf of Mexico Red Snapper, 1972-2011
More informationVeterinary Medical Terminology
Curriculum Outline: Course # Required courses prior to admission Credit hours BIO 0 Principles of Biology I with Lab 4 CHM 0 General Chemistry I with Lab 4 ENG 110 or 111 or 1 Freshman Composition or Composition
More informationAnswers to Questions about Smarter Balanced 2017 Test Results. March 27, 2018
Answers to Questions about Smarter Balanced Test Results March 27, 2018 Smarter Balanced Assessment Consortium, 2018 Table of Contents Table of Contents...1 Background...2 Jurisdictions included in Studies...2
More informationENGINEERING TEST SPECIFICATION
DATE PREPARED CREATED BY DATE ISED ISED BY LOGGED 1 of 6 1. PURPOSE To provide quality assurance during the manufacturing processes of the AuraSound NS3-194-8E full range speaker and to define the standard
More informationMammalogy Lecture 8 - Evolution of Ear Ossicles
Mammalogy Lecture 8 - Evolution of Ear Ossicles I. To begin, let s examine briefly the end point, that is, modern mammalian ears. Inner Ear The cochlea contains sensory cells for hearing and balance. -
More informationMOON PHASES FOR 2018, at Kitt Peak Times and dates are given in local time, zone = 7 hr West. They are generally better than +- 2 minutes.
Nighttime astronomical calendar program. Select a site: *SELECT SITE* - Enter single-character code: n.. new site (enter all parameters). x.. exit without change (current: Kitt Peak) k.. Kitt Peak s..
More informationMOON PHASES FOR 2019, at Kitt Peak Times and dates are given in local time, zone = 7 hr West. They are generally better than +- 2 minutes.
Nighttime astronomical calendar program. Select a site: *SELECT SITE* - Enter single-character code: n.. new site (enter all parameters). x.. exit without change (current: Kitt Peak) k.. Kitt Peak s..
More informationThe City School. Learn Create Program
Learn Create Program What is Scratch? Scratch is a free programmable toolkit that enables kids to create their own games, animated stories, and interactive art share their creations with one another over
More informationSubdomain Entry Vocabulary Modules Evaluation
Subdomain Entry Vocabulary Modules Evaluation Technical Report Vivien Petras August 11, 2000 Abstract: Subdomain entry vocabulary modules represent a way to provide a more specialized retrieval vocabulary
More informationX-bar Theory Motivating intermediate projections
X-bar Theory Motivating intermediate projections Ling 322 Read Syntax, Ch. 6 (Lecture notes based on Andrew Carnie s notes) 1 Flat structure (D) (Adj+) (+) (1) I bought [that big book of poems with the
More informationEVM analysis of an Interference Limited SIMO-SC System With Independent and Correlated Channels
EVM analysis of an Interference Limited SIMO-SC System With Independent and Correlated Channels arxiv:8.54v [cs.it] 26 Nov 28 Sudharsan Parthasarathy (a), Suman Kumar (b), Sheetal Kalyani (a) (a) Dept.
More informationGoal: To learn about the advantages and disadvantages of variations, by simulating birds with different types of beaks competing for various foods.
Name Date Activity: Bird Beak Adaptation Lab Goal: To learn about the advantages and disadvantages of variations, by simulating birds with different types of beaks competing for various foods. Background
More informationSimulation of the ASFA system in an ERTMS simulator
Computers in Railways XI 853 Simulation of the ASFA system in an ERTMS simulator I. Gómez-Rey, J. M. Mera & A. Lorenzo CITEF, Universidad Politécnica de Madrid, Spain Abstract Due to the economic, technical
More informationDesign of High Speed Vedic Multiplier Using Carry Select Adder with Brent Kung Adder
Design of High Speed Vedic Multiplier Using Carry Select Adder with Brent Kung Adder Kathi Anoosha M.Tech(VLSI&ES), AVN Institute of Engineering and Technology. Sasi Kiran, M.Tech Assistant Professor,
More informationHandling missing data in matched case-control studies using multiple imputation
Handling missing data in matched case-control studies using multiple imputation Shaun Seaman MRC Biostatistics Unit, Cambridge, UK Ruth Keogh Department of Medical Statistics London School of Hygiene and
More informationX-bar Node Flavors Introduction to syntax. Noun Phrase
X-bar Node Flavors Introduction to syntax Adam Szczegielniak Noun Phrase It can be replaced by a Pronoun: I think the cat likes [salty fish]k, and I think she eats [them]k often. It can be coordinated
More informationVirtual Genetics Lab (VGL)
Virtual Genetics Lab (VGL) Experimental Objective I. To use your knowledge of genetics to design and interpret crosses to figure out which allele of a gene has a dominant phenotype and which has a recessive
More informationMODELING THE CAUSES OF LEG DISORDERS IN FINISHER HERDS
ISAH-2007 Tartu, Estonia 417 MODELING THE CAUSES OF LEG DISORDERS IN FINISHER HERDS Birk Jensen, T., Kristensen, A.R. and Toft, N. Department of Large Animal Sciences, Faculty of Life Sciences, University
More informationA Column Generation Algorithm to Solve a Synchronized Log-Truck Scheduling Problem
A Column Generation Algorithm to Solve a Synchronized Log-Truck Scheduling Problem Odysseus 2012 Greg Rix 12 Louis-Martin Rousseau 12 Gilles Pesant 13 1 Interuniversity Research Centre on Enterprise Networks,
More informationDesign of Modified Low Power and High Speed Carry Select Adder Using Brent Kung Adder
Journal From the SelectedWorks of Kirat Pal Singh August, 2016 Design of Modified Low Power and High Speed Carry Select Adder Using Brent Kung Adder Amala Maria Alex, Mangalam college of Engineering, Kottayam,
More informationHuman Uniqueness. Human Uniqueness. Why are we so different? 12/6/2017. Four Candidates
Our Hominid Ancestors In humans, brain tissue has more than doubled over the past 2 million years. Break from chimps 3-5 million Our Hominid Ancestors Our Hominid Ancestors Relative Brain Size in Our Ancestors
More informationLABORATORY EXERCISE 7: CLADISTICS I
Biology 4415/5415 Evolution LABORATORY EXERCISE 7: CLADISTICS I Take a group of organisms. Let s use five: a lungfish, a frog, a crocodile, a flamingo, and a human. How to reconstruct their relationships?
More informationInformed search algorithms
Revised by Hankui Zhuo, March 12, 2018 Informed search algorithms Chapter 4 Chapter 4 1 Outline Best-first search A search Heuristics Hill-climbing Simulated annealing Genetic algorithms (briefly) Local
More informationPierre-Louis Toutain, Ecole Nationale Vétérinaire National veterinary School of Toulouse, France Wuhan 12/10/2015
Antimicrobial susceptibility testing for amoxicillin in pigs: the setting of the PK/PD cutoff value using population kinetic and Monte Carlo Simulation Pierre-Louis Toutain, Ecole Nationale Vétérinaire
More informationData Mining Classification: Alternative Techniques. Lecture Notes for Chapter 4. Rule-Based. Introduction to Data Mining, 2 nd Edition
Data Mining Classification: Alternative Techniques Lecture Notes for Chapter 4 RuleBased Introduction to Data Mining, 2 nd Edition by Tan, Steinbach, Karpatne, Kumar RuleBased Classifier Classify records
More informationChapter VII Non-linear SSI analysis of Structure-Isolated footings -soil system
Chapter VII 192 7.1. Introduction Chapter VII Non-linear SSI analysis of Structure-Isolated footings -soil system A program NLSSI-F has been developed, using FORTRAN, to conduct non-linear soilstructure
More informationLecture 11 Wednesday, September 19, 2012
Lecture 11 Wednesday, September 19, 2012 Phylogenetic tree (phylogeny) Darwin and classification: In the Origin, Darwin said that descent from a common ancestral species could explain why the Linnaean
More informationJune 2009 (website); September 2009 (Update) consent, informed consent, owner consent, risk, prognosis, communication, documentation, treatment
GUIDELINES Informed Owner Consent Approved by Council: June 10, 2009 Publication Date: June 2009 (website); September 2009 (Update) To Be Reviewed by: June 2014 Key Words: Related Topics: Legislative References:
More informationActivity 1: Changes in beak size populations in low precipitation
Darwin s Finches Lab Work individually or in groups of -3 at a computer Introduction The finches on Darwin and Wallace Islands feed on seeds produced by plants growing on these islands. There are three
More informationChapter 6: Extending Theory
L322 Syntax Chapter 6: Extending Theory Linguistics 322 1. Determiner Phrase A. C. talks about the hypothesis that all non-heads must be phrases. I agree with him here. B. I have already introduced D (and
More informationAn Esterel Virtual Machine (EVM) Aruchunan Vaseekaran
An Esterel Virtual Machine (EVM) Aruchunan Vaseekaran Why Esterel is suited for Deterministic Control Systems Imperative Language Synchronous Concurrency, Preemption Not widely available in low cost systems.
More informationINTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN ENGINEERING AND TECHNOLOGY (IJARET)
INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN ENGINEERING AND TECHNOLOGY (IJARET) International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 ISSN 0976-6480 (Print) ISSN
More informationMathematical models for dog rabies that include the curtailing effect of human intervention
Mathematical models for dog rabies that include the curtailing effect of human intervention Tiffany Ngo Leung Supervised by Dr Stephen A Davis RMIT University Abstract Rabies is a zoonotic viral disease
More informationCow Exercise 1 Answer Key
Name Cow Exercise 1 Key Goal In this exercise, you will use StarGenetics, a software tool that simulates mating experiments, to analyze the nature and mode of inheritance of specific genetic traits. Learning
More informationDesign of Low Power and High Speed Carry Select Adder Using Brent Kung Adder
Design of Low Power and High Speed Carry Select Adder Using Brent Kung Adder Dr.K.Srinivasulu Professor, Dept of ECE, Malla Reddy Collage of Engineering. Abstract: The binary addition is the basic arithmetic
More informationPage # Events Page # Previous Event/Explanation 4 Kitten tried to lick the moon and she got a bug on her tongue. milk.
START:READ 1 Guide for Repeated Interactive Read-Alouds Complex Storybook Goals for Read 1 Kitten s First Full Moon by Kevin Henkes Push In and Connect Key Events Push-In Story Problem and Target Vocabulary
More informationINFO 1103 Homework Project 1
INFO 1103 Homework Project 1 January 22, 2018 Due February 7, at the end of the lecture period. 1 Introduction Many people enjoy dog shows. In this homework, you will focus on modelling the data represented
More informationComparison of Parallel Prefix Adders Performance in an FPGA
International Journal of Engineering Research and Development e-issn: 2278-067X, p-issn: 2278-800X, www.ijerd.com Volume 3, Issue 6 (September 2012), PP. 62-67 Comparison of Parallel Prefix Adders Performance
More informationPurpose and focus of the module: Poultry Definition Domestication Classification. Basic Anatomy & Physiology
Module: Poultry Production Code: AP21 Purpose and focus of the module: It aims at providing students with adequate knowledge and skills in poultry husbandry techniques and farm management. Skill Objectives
More informationEntailment above the word level in distributional semantics
Entailment above the word level in distributional semantics Marco Baroni Raffaella Bernardi Ngoc-Quynh Do Chung-chieh Shan University of Trento University of Trento EM LCT, Free University of Bozen-Bolzano
More informationTurtle Ballet: Simulating Parallel Turtles in a Nonparallel LOGO Version. Erich Neuwirth
Turtle Ballet: Simulating Parallel Turtles in a Nonparallel LOGO Version Erich Neuwirth University of Vienna, Dept. of Statistics and Decision Support Systems Computer Supported Didactics Working Group
More informationLevel 5. Book a. Level 5. Word Count 98 Text Type Narrative High Frequency Word/s Introduced. The Thirsty Cats. Working Dogs.
Level 5 Book a The Thirsty Cats Pet Dogs and Working Dogs This Little Critter To the Space Station Big Greg, the Firefighter Level 5 Word Count 98 Text Type Narrative High Frequency Word/s Introduced Trent
More informationImproving consumer protection against zoonotic diseases Phase II Project No: EuropeAid/133990/C/SER/AL
Indicative Budget for implementation of the Bovine Brucellosis Control Programme in 2016 1. Introduction ANNEX 6.1 Given the current epidemiological situation of brucellosis in cattle in Albania, a flexible,
More informationWe recommend you cite the published version. The publisher s URL is
Adamatzky, A. (2010) Physarum machines: computers from slime mould. (74) World Scientific Publishing. ISBN 978-981-4327-58-9 We recommend you cite the published version. The publisher s URL is Refereed:
More informationVeterinarian Assistant
Statewide Framework Document for: 510808 Standards may be added to this document prior to submission, but may not be removed from the framework to meet state credit equivalency requirements. Performance
More informationChapter 6: Extending Theory
&L322 Syntax Chapter 6: Extending Theory Linguistics 322 1. Determiner Phrase A. C. talks about the hypothesis that all non-heads must be phrases. I agree with him here. B. I have already introduced D
More informationDevelopment of the New Zealand strategy for local eradication of tuberculosis from wildlife and livestock
Livingstone et al. New Zealand Veterinary Journal http://dx.doi.org/*** S1 Development of the New Zealand strategy for local eradication of tuberculosis from wildlife and livestock PG Livingstone* 1, N
More informationTREAT Steward. Antimicrobial Stewardship software with personalized decision support
TREAT Steward TM Antimicrobial Stewardship software with personalized decision support ANTIMICROBIAL STEWARDSHIP - Interdisciplinary actions to improve patient care Quality Assurance The aim of antimicrobial
More informationThe Kaggle Competitions: An Introduction to CAMCOS Fall 2015
The Kaggle Competitions: An Introduction to CAMCOS Fall 15 Guangliang Chen Math/Stats Colloquium San Jose State University August 6, 15 Outline Introduction to Kaggle Description of projects Summary Guangliang
More informationCh 1.2 Determining How Species Are Related.notebook February 06, 2018
Name 3 "Big Ideas" from our last notebook lecture: * * * 1 WDYR? Of the following organisms, which is the closest relative of the "Snowy Owl" (Bubo scandiacus)? a) barn owl (Tyto alba) b) saw whet owl
More informationIntroduction to Python Dictionaries
Introduction to Python Dictionaries Mar 10, 2016 CSCI 0931 - Intro. to Comp. for the Humanities and Social Sciences 1 ACT2-4 Let s talk about Task 2 CSCI 0931 - Intro. to Comp. for the Humanities and Social
More informationAccounting for the causal link between free adjuncts and their host clauses
Accounting for the causal link between free adjuncts and their host clauses Sinn und Bedeutung 23, Barcelona Sarah Zobel September 7, 2018 Slides at: http://sarahzobel.net Introducing free adjuncts I Free
More informationCONTINUING EDUCATION AND INCORPORATION OF THE ONE HEALTH CONCEPT
CONTINUING EDUCATION AND INCORPORATION OF THE ONE HEALTH CONCEPT M. Farnham 1, W. Hueston 2 Original: English Summary: Sixteen Members of the OIE Regional Commission for the Middle East responded to a
More informationBioSci 110, Fall 08 Exam 2
1. is the cell division process that results in the production of a. mitosis; 2 gametes b. meiosis; 2 gametes c. meiosis; 2 somatic (body) cells d. mitosis; 4 somatic (body) cells e. *meiosis; 4 gametes
More informationGenetics Lab #4: Review of Mendelian Genetics
Genetics Lab #4: Review of Mendelian Genetics Objectives In today s lab you will explore some of the simpler principles of Mendelian genetics using a computer program called CATLAB. By the end of this
More informationFrom James Merrill's Dogs to the. Alfred A. Knopf Borzoi Devices. Jack W. C. Hagstrom
From James Merrill's Dogs to the Alfred A. Knopf Borzoi Devices Jack W. C. Hagstrom by Jack W. C. Hagstrom - page 1 In the process of describing the poet James Merrill's books published by Alfred A. Knopf
More informationLab 7. Evolution Lab. Name: General Introduction:
Lab 7 Name: Evolution Lab OBJECTIVES: Help you develop an understanding of important factors that affect evolution of a species. Demonstrate important biological and environmental selection factors that
More informationHeuristic search, A* CS171, Winter 2018 Introduction to Artificial Intelligence Prof. Richard Lathrop. Reading: R&N
Heuristic search, A* CS171, Winter 2018 Introduction to Artificial Intelligence Prof. Richard Lathrop Reading: R&N 3.5-3.7 Outline Review limitations of uninformed search methods Informed (or heuristic)
More information[EMC Publishing Note: In this document: CAT 1 stands for the C est à toi! Level One Second Edition Teacher s Annotated Edition of the Textbook.
EMC Publishing s Correlation of C est à toi! Levels One, Two, Three 2 nd edition to the 2007 Indiana Academic Standards for World Languages 9-12 Sequence - Modern European and Classical Languages Grade
More informationModeling and Control of Trawl Systems
Modeling and Control of Trawl Systems Karl-Johan Reite, SINTEF Fisheries and Aquaculture Supervisor: Professor A. J. Sørensen * Advisor: Professor H. Ellingsen * * Norwegian University of Science and Technology
More informationImplementation and Estimation of Delay, Power and Area for Parallel Prefix Adders
International Journal for Modern Trends in Science and Technology Volume: 02, Issue No: 11, November 2016 http://www.ijmtst.com ISSN: 2455-3778 Implementation and Estimation of Delay, Power and Area for
More informationPerformance Analysis of HOM in LTE Small Cell
Sensors & Transducers, Vol. 170, Issue 5, May 014, pp. 160-164 Sensors & Transducers 014 by IFSA Publishing, S. L. http://www.sensorsportal.com Performance Analysis of HOM in LTE Small Cell Juan HANG Practice
More informationST NICHOLAS COLLEGE HALF YEARLY PRIMARY EXAMINATIONS. February YEAR 4 ENGLISH TIME: 1hr 15 min (Reading Comprehension, Language, and Writing)
ST NICHOLAS COLLEGE HALF YEARLY PRIMARY EXAMINATIONS February 2016 YEAR 4 ENGLISH TIME: 1hr 15 min (Reading Comprehension, Language, and Writing) Marking Scheme A. Reading Comprehension (20 marks) 1. Tick
More information10015NAT Graduate Diploma Veterinary Acupuncture
10015NAT Graduate Diploma Veterinary Acupuncture Nationally accredited under the Australian Qualifications Framework at postgraduate level. WHO IS IT FOR? The 10015NAT Graduate Diploma Veterinary Acupuncture
More information