CSci 6360
Data Science II

John Miller
Spring 2023

Textbook

Textbook (ScalaTion): Introduction to Computational Data Science Using ScalaTion, 2020.
John A. Miller
Introduction to Data Science Using ScalaTion: Lesson Plans, 2020.
John A. Miller

Textbook (ISL): An Introduction to Statistical Learning, 2013.
Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani
TextBook (ESL): The Elements of Statistical Learning, 2nd Ed., 2009.
Trevor Hastiei, Robert Tibshirani and Jerome Friedman
TextBook (MML): Mathematics for Machine Learning 2020.
Marc Peter Deisenroth, A Aldo Faisal and Cheng Soon Ong

Guides

R: Linear Regression Using R: An Introduction to Data Modeling, 2016.
David J. Lilja
Spark: Machine Learning Library (MLlib) Guide, 2018.
Apache Spark

Class Time

Day Time Plan Room

Tuesday 12:45-2:00 Lecture + TT Miller Plant Science (1061) 2102
Wednesday 12:40-1:30 Lecture Boyd (1023) 208
Thursday 12:45-2:00 Lecture + HW Miller Plant Science (1061) 2102

Course Description

An introduction to advanced analytics techniques in data science, including random forests, semi-supervised learning, spectral analytics, randomized algorithms, and just-in-time compilers. Distributed and out-of-core processing.

Course Topics

ScalaTion Chapters:

Part I Preliminaries

Ch. 1: Introduction to Data Science
Ch. 2: Linear Algebra
Ch. 3: Probability
Ch. 4: Data Management
Ch. 5: Data Preprocessing
Part II Modeling
Ch. 6: Prediction
Ch. 7: Classification
Ch. 8: Classification - Continuous Variables
Ch. 9: Generalized Linear Models and Regression Trees
Ch. 10: NonLinear Models and Neural Networks
Ch. 11: Time-Series/Temporal Models
Ch, 12: Multivariate and Nonlinear Time-Series
Ch. 13: Dimensionality Reduction
Ch. 14: Clustering
Part III Simulation
Ch. 16: Simulation Foundations
Ch. 16: State-Space Models
Ch. 17: Event-Oriented Models
Ch. 18: Process-Oriented-Space Models
Ch. 19: Simulation Output Analysis
Part IV Appendices
A. Optimization in Data Science
B. Functional Data Analysis
C. Graph Databases and Analytics

ISL Chapters:

Ch. 1: Introduction
Ch. 2: Statistical Learning
Ch. 3: Linear Regression
Ch. 4: Classification
Ch. 5: Resampling Methods
Ch. 6: Linear Model Selection and Regularization
Ch. 7: Moving Beyond Linearity
Ch. 8: Tree-Based Methods
Ch. 9: Support Vector Machines
Ch. 10: Unsupervised Learning

ESL Chapters:

Ch. 1: Introduction
Ch. 2: Overview of Supervised Learning
Ch. 3: Linear Methods for Regression
Ch. 4: Linear Methods for Classification
Ch. 5: Basis Expansions and Regularization
Ch. 6: Kernel Smoothing Methods
Ch. 7: Model Assessment and Selection
Ch. 8: Model Inference and Averaging
Ch. 9: Additive Models, Trees, and Related Methods
Ch. 10: Boosting and Additive Trees
Ch. 11: Neural Network
Ch. 12: Support Vector Machines and Flexible Discriminants
Ch. 13: Prototype Methods and Nearest-Neighbors
Ch. 14: Unsupervised Learning
Ch. 15: Random Forests
Ch. 16: Ensemble Learning
Ch. 17: Undirected Graphical Models
Ch. 18: High-Dimensional Problems

Grading

20% Exam I 2/?
20% Exam II 4/?
25% Final Exam 5/?
30% Programs (groups of 3-4 students) see Software
5% Homework/Tool Talks [presentation]

Software:

Exam I
Review Date: Feb ?, 2023
Exam Date: Feb ?, 2023
5 Questions:

Exam II
Review Date: Apr ?, 2023
Exam Date: Apr ?, 2023
5 Questions:

Final Exam
Exam Date: May ?, 2023
6 Questions:

HW: Homework (pair)

Requirement: Present 1 (pair of students).

See eLC.

TT: Tool Talks (group)

Requirement: Present 1 (group).

See the TAs Updated List.

No. Topic Description Talk Due

1. R R Language . .
2. Spark Apache Spark . .
3. MLlib Apache Spark MLlib . .
4. TensorFlow TensorFlow Machine Learning Library . .
5. Keras The Python Deep Learning Library . .

6. Weka Waikato Environment for Knowledge Analysis (Weka) . .
7. Watson, Cognos Analytics Watson Analytics/Cognos Analytics . .
8. Parallel Matrix Operations GPU/TPU/FPGA . .

Projects

TA:

Submit projects to the TA following their instructions (one submission per group ). Turn in your fully commented source code files, an SBT build.sbt file and a ReadMe file. The ReadMe file must contain instructions for compiling and running the program as well as a detailed explanation of who coded what parts of the program. Five percent of the grade is determined by how well the project is documented. Another five percent of the grade is determined by how well the effort is divided between the group members.

See eLC for the three regular projects and more details on the term project. Project #1 on Ch. 6, Project #2 on Ch. 10, Project #3 TBD.

No. Description Starter Code (must be used) Comment Due

4. Term Project: Data Science Application . A two-page proposal giving a detailed description of the application you propose to develop must be submitted with Project 2. Project includes data collection, data analytics, interpretation and recommendations for a real-world project. May use ScalaTion, R, Scikit Learn, Keras, Spark or other approved software. The term project including a 25-minute presentation and demo will be presented during the last week of class. Must address the ten points/questions listed below. Worth twice the points of regular projects. .

Ten Questions:

problem statement (focus/purpose of study)
describe datasets (at least 2)
data preprossing (techniques applied)
visual examination of data
modeling techniques used (at least 3)
why they were chosen
quality of fit for model, parameters
feature selection
interpretation of results
recommendations of study

Optional Project:

Elastic Net Regression: f_opt = sse + lambda ((1-alpha) b.normSq + alpha b.norm1). When alpha = 1 it becomes Lasso Regression and when alpha = 0 it becomes Ridge Regression. Between 0 and 1, it is a hybrid. Need to finish other factorization for Ridge Regression and implement Elastic Net Regression.
Implement Gaussian TAN Bayes (TANBayesR) that supports continuous variables/features.
Implement Hidden Markov Models (HMM).
Implement Partial Least Squares (PLS).
Select a modeling technique and get approval.

Policies

Late Points - 10 points off per day late.

Make-Up Tests - written pre-approval or medical documentation required (no retakes).

A Culture of Honesty -- Examples.
Copyright Issues -- Regents Guide to Understanding Copyright.

Day	Time	Plan	Room
Tuesday	12:45-2:00	Lecture + TT	Miller Plant Science (1061) 2102
Wednesday	12:40-1:30	Lecture	Boyd (1023) 208
Thursday	12:45-2:00	Lecture + HW	Miller Plant Science (1061) 2102

20%	Exam I	2/?
20%	Exam II	4/?
25%	Final Exam	5/?
30%	Programs (groups of 3-4 students)	see Software
5%	Homework/Tool Talks	[presentation]

No.	Topic	Description	Talk	Due
1.	R	R Language	.	.
2.	Spark	Apache Spark	.	.
3.	MLlib	Apache Spark MLlib	.	.
4.	TensorFlow	TensorFlow Machine Learning Library	.	.
5.	Keras	The Python Deep Learning Library	.	.
6.	Weka	Waikato Environment for Knowledge Analysis (Weka)	.	.
7.	Watson, Cognos Analytics	Watson Analytics/Cognos Analytics	.	.
8.	Parallel Matrix Operations	GPU/TPU/FPGA	.	.

No.	Description	Starter Code (must be used)	Comment	Due
4.	Term Project: Data Science Application	.	A two-page proposal giving a detailed description of the application you propose to develop must be submitted with Project 2. Project includes data collection, data analytics, interpretation and recommendations for a real-world project. May use ScalaTion, R, Scikit Learn, Keras, Spark or other approved software. The term project including a 25-minute presentation and demo will be presented during the last week of class. Must address the ten points/questions listed below. Worth twice the points of regular projects.	.

CSci 6360 Data Science II

John Miller Spring 2023