Database Systems - Florida Institute of Technology



Artificial Intelligence

CSE 5290/4301, Fall 2019

Instructor: Debasis Mitra, Ph.D.

Office: Harris 325 E-mail: dmitra ‘at’ cs.fit. edu

Class Home Page:

|Class Room: Olin Physical Sci 144 |Time: MW 1:00-2:15 pm |

Office Hours: MW 10am-12 pm (or by appointment)

Tentative Grading plan:

Graduate stds: Quizzes (6): 15%, Coding exercise: 5%, Mid Term: 15%, Final Exam: 25%, Project: 40%

Undergraduate stds: Quizzes (6): 20%, Coding exercises: 30%, Mid Term 20%, Final Exam: 30%

SYLLABUS FOR AI FALL 2019 FINAL EXAM.

*Materials under *---* are excluded. 3.3-3.4.5 means 3.3 through 3.4.5

SEARCH

Ch 3.3-3.4.5, *bi-directional search*,3.4.7-3.5.2, *memory bounded srch*, 3.6-3.6.3

Ch 4-4.2

Ch 5-5.3

CONSTRAINTS

Ch 6-6.2.5, 6.3-6.3.2, 6.5 (*use of symmetries on p226*)

*Temporal reasoning: My slides*

LOGIC

Ch 7.3-7.5.2, *completeness of resolution*, 7.5.3-7.6.2, *walksat algorithm*,

Ch 8.2-8.3

Ch 9-9.2.2, 9.3-9.3.2, 9.4.1, 9.5-9.5.3

PROBABILISTIC REASONING

Ch 13.2-end

Ch 14-14.3

LEARNING

Decision Tree 18 – 18.3.4

Evaluation 18.4-Model Selection 18.4.1

Regularization 18.4.3

Learning Theory 18.5.0 only

Regression 18.6 – 18.6.2

Classification 18.6.3 – 18.6.4

Neural Network 18.7 – 18.7.4 (exclude exotic varieties of NN on my slides)

Non-parametric models 18.8 – 18.8.4

SVM basics 18.9

Clustering basics (from my slides)

ETHICS

My slides

Florida Tech Academic Calendar:

Detailed activities Fall 2019 mapped on to the dates in Spring 2019, also acts as our developing plan for this semester: ~ 28 meetings

Lectures planned: Search 4, Constraints 3, Auto Reasoning 4, Prob Reasoning 4, Machine Learning 5, Ethics 1,

Exams 3, Std presentations 4

|Date |Activities planned or performed last Spring | |

|Aug 19, M |AI an Introduction, and get-to-know quiz |My thoughts on Grad projects for now |

| | | |

|Aug 21, W |Required CS background list | |

| |Big O-notation (Algo-Intro slide 1-10), | |

| |Djikstra algorithm (Algo-Graph slide 23-25), Min-spanning tree | |

| |algorithm (Algo-Graph slide 54, 58) | |

| | | |

| |AI SEARCH (From my slides): 8-puzzle: | |

| | | |

| |BFS- Uniform Cost; DFS-Depth Limited | |

|Aug 26, M |AI SEARCH (From my slides): 8-puzzle: |Quiz-1 due |

| | |Depth Limited (DL): Fixed depth l as input parameter, |

| |BFS- Uniform Cost; DFS-Depth Limited (up to sl 22) | |

| |NP-completeness (Algo-Complexity slides 6-9:NP; 10-14: Problem class;|Iterative Deepening: |

| |15-22: PvsNP; 32-33: CLK-thm; |DL with increasing l=2, 3, … until goal is found. |

| |41-46: NP-complete; 63-67: What-to-do-with-NPcomplete) | |

| | | |

| |Iterative-deepening, | |

| | | |

| |Coding exc 1 Announced (below here) | |

| |Due 1/22/T/Class/Hard copy submission | |

| | | |

| |Graduate Projects Announced | |

|Aug 28, W |Heuristic Search A*, IDA*, SMA* |Graduate projects: /Fall2019/GradProjectLists.doc |

| |MySlides 15-35 | |

| | |Coding Exercise-1 (below this table) due 9/4/W |

| |NP-completeness (Algo-Complexity slides: Problem class; 15-22: PvsNP;| |

| |32-33: CLK-thm; | |

| |41-46: NP-complete; | |

|Sep 2, M |-- Labor day --- | |

| |UG Project-1 announced, | |

| |due Feb 19/T/InClass/HardCopy | |

|Sep 4, W |-- Hurricane Dorian day -- | |

|Sep 9, M |Grad projects – more detail, Proposal due: 9/16/M |Code-1 due. |

| |Algo slide 63-67: What-to-do-with-NPcomplete) | |

| |Text Slide Ch4a | |

|Sep 11, W |My SearchSlides (31) | |

| |Local search from Text-slides Ch4b: | |

| |Game (Adversarial) search | |

| |Local search, Hill-climbing, Simulated Annealing, Local beam search, | |

| |Genetic Algorithm, Gradient search, | |

| |MySlide (36-end) | |

|Sep 16, M |Remind imp of heuristics in PRUNING search tree; |Grad Project written proposal due in class, |

| |TextSlide Ch4b- sl7 to explain Gradient Search on continuous space |hard copy 2-3 pages (any format is ok) |

| |sl13; | |

| |MySlides (search): sl 44-47 (exclude sl48-end) |Next UG assignment-2: |

| | |Code IDA* on Romania road network |

| |REASONING WITH CONSTRAINTS: Motivating with Map/Graph coloring, |Due: ?? (on or after Oct 7) |

| |Backtracking | |

| |Backtracking, Forward Checking (TextSlides) | |

|Sep 18, W |Ch5 text slides 32-end, then, | Grad Project written proposal feedback |

| |MySlides, then Animated one; AC-3, PC-2, SAT-DPLL, WalkSat | |

| | | |

| |SPATIO-TEMPORAL CONSTRAINT REASONING from my slides. | |

| |A relevant web page: | |

| | | |

|Sep 23, M |AUTOMATED REASONING (ch 7): Syntax-Semantics-Model, |Undergrad Coding assignment-2 on IDA* search, |

| |Satisfaction-Entailment-Inference procedure-Validity; |at the bottom of this table. Ask for any clarification. |

| | |Due: October 16 in class |

|Sep 25, W |Brief intro to artificial neural network: My slides 23-4, 32, 34, 43|Grad Project updated written proposal. |

| | | |

| |AUTOMATED REASONING – continued (from slide 32): | |

| |Propositional Knowledge Base, Model checking algorithm, | |

| |Forward chaining algo, Backward chaining algo (up to sl#66) | |

|Sep 30, M |Quiz on Search and Constraint reasoning |A sentence not in Horn form: P is not true, ~P=>T |

| | | |

| |CNF, Resolution Algo (p255), Horn Clause, Definite clause | |

| |AUTOMATED REASONING: First Order Logic-Motivation; | |

| |Model-Interpretation-Quantifiers-Inferencing | |

| |Completeness-Herbrand Universe, Resolution strategies | |

| |Unification, Forward Chaining, Backward Chaining, | |

|Oct 2, W |Explain Horn clause, definite clause, in CNF | |

| |AUTOMATED REASONING: inferencing algorithms | |

|Oct 7, M |Possible spill-over from last class: sl24, then sl17-end |Grad Project First Progress Report due: update proposal, |

|(Mid term grades |Graduate Project presentation: ~10 min each group |include work that has been done so far; |

|open 10/4) |(presentation schedule posted on project-tracker doc, come early and | |

| |make sure you can project properly: | |

| |slides defining the problem, any result, demo, …) | |

|Oct 9, W |Possible spill-over from presentations |Feedbacks on graduate students’ progress reports. |

| |Return papers | |

| |?Discuss some exc problems | |

| | | |

| |A short online quiz to test canvas accessibility, bring laptop or | |

| |smart phone | |

|Oct 14-15, MW |-- Fall break -- | |

|Oct 16, W |Resolution – last few slides |Undergrad coding assignment-2 on IDA* search due: extended to Oct 23 in class. |

| | | |

| |MODELING UNCERTAINTY, Ch 14: Motivation | |

| |Probability-Joint-Inference-Conditional-Baye’s rule; | |

|Oct 21, M |REASONING WITH UNCERTAINTY: | |

| |Bayesian Net Probability reasoning contd. | |

| |Reasoning with probability, Ch 13: Sl 10-26 | |

|Oct 23, W |Reasoning with probability, Ch 13: Sl 27-end |Grad stds: Schedule meeting during office hours |

| |Bayesian network Ch 14-start |to discuss your project if necessary, those who cannot come during my office hrs may|

| | |meet me after the class |

| |Exercise problems done in class | |

| |Undergrad coding assignment-2 on IDA* search due | |

|Oct 28, M |Bayesian network Ch 14-finish | MidTerm: some questions may be too long to view on smartphone. |

| | | |

| | | |

| |Mid Term: On line Canvas - bring your laptop or smart phone for the | |

| |exam! Multiple-choice. | |

| |In class, 30-40 min, Closed book. | |

| |Syllabus: Search, Constraint, Logic. | |

|Oct 30, W |-- at conference -- | |

|Nov 4, M |Ch 13 exercise | |

| |Sample questions Ch13: 13.8, 14, 15 | |

| | | |

| |Grad Project presentation: ~10 min each group | |

|Nov 6, W |Grad Project presentation continued: ~10 min each group | |

| |MAKE UP CLASS ON SATURDAY, same time-place as that of the regular | |

| |class | |

| | | |

| |MACHINE LEARNING: Decision tree, basics: up to slides 27 today | |

| | | |

| |15 minute online test on Probabilistic reasoning (Ch 13, 14) | |

|Nov 9, Sa (make-up|Ch 18.6: MACHINE LEARNING continued: MDL, PAC, | |

|class) |18.6.1-2. Regression: Linear, | |

| |Multi-variate regression, linear | |

|Nov 11, M |-- Veterans day -- | |

|Nov 13, W |18.6.3-4. Classifiers, Logistic regression, Perceptron 18.7.1-4. | |

| |Artificial-neural Networks [Start from MySlide#24, exclude slides | |

| |39-43] | |

| |18.7.1-4. Artificial-neural Networks | |

| |Ch 18.8: non-parametric, kNN and variations | |

|Nov 18, M |Ch 18.8: non-parametric: LSH (start - slide 51), kernel-regression |Graduate Project Final report due Monday 18th by 9pm |

| |and 18.9-10 SVM |by e-mail (i) A report between 4-10 pages, including |

| |Clustering basics: K-means, from |an abstract and references in NIPS conference format |

| | |(), you need not use |

| | |LaTex, but just follow the formatting style., |

| | |(ii) Your source code including |

| | |instructions on how to run and any dependencies and |

| | |library details, |

| | |and (iii) data files with metadata information (to understand |

| | |data). |

| | |All three in a zip file, named using |

| | |Sp19 _project name’s signature id_one group member |

| | |Send by e-mail. |

|Nov 20, W |Clustering: K-median, Slide 64 hierarchical, density-based | Comments on grad-project final reports may be below. |

| | | |

| |AI and Ethics (my slides are included in syllabus/tests) | |

| | | |

| |Online (make-up) quiz on probabilistic reasoning: 15 min |UPDATED GRAD-PROJECT REPORTS ARE DUE SUNDAY EVENING BY E-MAIL, |

| | |ONLY THE REPORT DOCUMENT |

|Nov 25, M |Decision tree revision |DISCLAIMER: CANVAS HAS NO FORMULA NOW. |

| | |I USE MY OWN SPREADSHEET. |

| |Hard-copy test on probabilistic reasoning: 20 min |YOU WILL SEE YOUR COMPREHENSIVE GRADE THERE IN A SEPARATE COLUMN IN CANVAS AND |

| | |LETTER GRADE IN PAWS. |

|Nov 27-29, W-Sun |-- Thanksgiving break -- |Two name-less papers from the last test (Bhumi & Mahbuba), I do not know which one |

| | |is whom! |

|Dec 2, M |Test on Machine Learning |We will have a final exam on its due time as below. |

| |Final exam discussion |Final: |

| |Do not forget the instructor-review! |You must bring your laptop or smart-phone |

| | |to answer multiple-choice |

| | |questions MCQ on Canvas (a part of the test). |

| | |Four problem questions for writing on answer paper, |

| | |and one MCQ set of questions on Canvas (12 minutes) |

| | |Each question 10 points. |

|Dec 4, W |(there will be no class) |You are responsible for looking exam time up on the FIT site for correct date. |

|(Dec 6 last class | |I provide the information here to help you only. |

|day) | | |

| |Part written questions like those in take-home / quizzes. Part online| |

| |canvas, like that in the Midterm, for a fixed time. Nineteen online | |

| |questions and ¾ offline questions, respectively. A subset of | |

| |questions are for Undergrads. Using a calculator is ok but showing | |

| |computation steps are most important. Showing only final results on | |

| |offline answers are NOT acceptable. | |

| | | |

| |Online access other than the test are strictly prohibited. | |

| | | |

| |Aggregate formula: Project 35%, Exam1 10%, Exam2 12%, Final 20%, | |

| |MidTerm 12%, Coding 11% | |

|FINAL EXAM: | up day classes, MW, in that web site |

| |schedules/fall-final-examination-schedule/ |Fall 2019 Grade distributions below. |

|1–1:50 p.m. MW |Tuesday, Dec. 10 |1–3 p.m. |

Grading: All columns within a group has equal weights. Below is the percentage on each group.

Undergrad: Quiz & tests (except ML): 20%; Codes: 30%; Midterms (including ML): 20%; Final: 30% A/B: 74 B/C: 65

Grad: Quiz & tests (except ML): 15%; Prog: 5%; Midterms (including ML): 15%; Final: 25%; Project: 40%

A/B: 85

I keep papers for a semester.

HAPPY HOLIDAYS!

COMMENTS ON FINAL REPORT (UPDATED REPORTS ARE DUE SUNDAY EVENING BY E-MAIL, ONLY THE REPORT DOCUMENT)

(Individual grading between partners may be done)

Maze-decision problem:

Fig 9: what are x-y axes?

Activation maps of success-failure inferencing over positive-negative samples?

Future: how to vary density? size? arbitrary start-end?

SOM on time-series:

Report file name as |”SOM resources.docx”?

What is the role of each partner? How much did Mahbuba learn?

Reference for choosing grid size?

Think more on observation (two members together) and find ways to quantitate such parameters: 1) on each image, and 2) comparing images.

Write references properly.

ANN-SAT:

Send me your doc or latex file. I may go for a conference submission by the weekend.

What are your ideas of future improvements, jot down a few.

Sensitivity test on features? < if you can do it, good, but at least discuss in the future works>

Topological clustering:

Refer each figure within the text.

Point to the references within the text.

Any conclusion on your results?

Motif recognition on IVC artifacts:

Before getting to data augmentation (section 3), talk on your data (section 2). Where-what-how many, etc.

Give specifics of augmentation. How many rotations, zoom, etc. Total augmented data for each original data.

How you label data? You should mention the objective of the project in your introduction of the report. Labels are according to that objective.

Before “output” section, talk on your experiment design. Training data size, cross-validation k=?, separate validation data, etc.

Provide Fig number, Table number for each and discuss them in the text.

Clearly, overfitting is happening (Fig above section 7). How should one address that? Discuss in the conclusion section.

Sec 8.2.3, connect to sec 7.4, on where are those accuracies coming from. I suggest, merge 8.2 with sec 7. Let 8.1 your learning be an independent section.

CODING EXERCISES

1. All students, individual assignment: DFS and BFS blind search. Use a binary tree as input: layers/rows are indicated with A, B, C, D, E; and nodes are indexed with 1, 2, 3…. in each layer/row. Submit output of your search printing the node name each time the control arrives at a node. Thus, only for two layers A and B, the output for DFS may be B1, B2, A1, and for BFS, A1, B1, B2. Indenting is a good way to make the output from search algorithms legible.

A non-CS student may partner with a CS student, but the submission must be individual. Both the individuals should mention the partners’ names.

Due: 9/4/2019/W/In-class, hard copy showing your code in any language (and yes, commented for me to understand) and output snapshots from the running code with CPU-time. [10 pts]

2. Only undergraduate Students (individual) assignment 2: Develop and run the IDA* algorithm on the Romanian road-network map from the textbook (Fig. 3.2, p68) for the start node as Arad and the goal node as Bucharest. Submission in hard copy: a) Source codes; b) Full search tree from running your code highlighting the shortest path to the goal. [30 pts]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download