Curriculum:

The Master of Science in Analytics degree offered by the College of Science is an inter-collegiate curriculum with the Mays Business School. A student seeking the 36 semester credit hours Master of Science Analytics degree must fulfill the requirements listed below. One statistics course completed with a C or  better is the prerequisite for the degree. Examples of acceptable courses at Texas A&M University include STAT 301, STAT 302 or STAT 303 or
INFO 303 or STAT 651 or equivalent.


A 5 Semester Part-Time Program

 

Fall Semester (Year one) 

Orientation at Houston CityCentre, Houston, TX
1. Technology overview
2. Introduction to teams
3. Introduction to programming and logic
4. Statistics refresher
5. Tableau, JMP & SAS training
6. Conflict management assessment
7. Project Management: Student panel on capstone
8. Program Overview

Prepare data for model fitting, fit appropriate regression models to business data from a wide variety of settings, interpret the output from regression models, identify weaknesses in models and formulate ways to overcome them, make valid inferences and draw business conclusions on the basis of the fitted models, and recommend business actions on the basis of these conclusions.

Topics covered in STAT 608
1. Regression methods based on least squares

2. Diagnostic methods including marginal model plots

3. Transformations and weighted least squares

4. Shrinkage and model selection techniques including lasso

5. Linear regression splines including MARS

6. Logistic regression

7. Poisson regression

8. Regression models with serially correlated errors

9. Bayesian approaches to linear and logistic regression

In STAT 608, students undertake a major project focused on building and interpreting one or more predictive models. Previous projects have focused on

1. Modeling the asking price of certified pre-owned BMW and Lexus automobiles

2. Modeling default probability in 30 year mortgages for single family homes using the Freddie Mac data

3. Modeling price and customer ratings of Airbnb apartment stays in New York City, Paris and San Francisco

4. Modeling price and popularity of New York City restaurants

5. Modeling property assessment values for condominium units across different wards in Boston

The primary objective of the course is to familiarize students with general database concepts, database design methodologies, and database implementation. In addition, the course introduces students to Hadoop as a distributed data storage technology used by companies to process large data sets and to Python as a computer coding language used to obtain data from social media.

Topics:
1. Introduction to relational databases and SQL
2. How to retrieve data from two or more tables
3. How to code summary queries
4. How to insert, update, and delete data
5. How to work with data types
6. How to work with functions
7. How to work with the views
8. Introduction to coding and coding standards and Python
9. Core objects, variables, input, and output in Python
10. Structures that control flow in Python
11. Functions in Python
12.Processing data in Python
13. Miscellaneous topics in Python

The student will learn:
1. Teamwork and team effectiveness
2. Understanding Big Data
3. Ethics
4. Project Management

Spring Semester (Year one) 

This course introduces foundations of multivariate analysis including matrix algebra, random vectors, multivariate distributions, and statistical inference for multivariate data. It also introduces methods of machine learning including principal component analysis, discriminant analysis, clustering, and elements of factor analysis and canonical analysis.

Topics covered:
1. Graphical techniques
2. Principal component analysis
3. Cluster analysis
4. Correspondence analysis and other ordination techniques
5. Discriminant analysis
6. Classification and prediction
7. Factor analysis and SEM
8. Inference on mean vectors
9. Canonical correlation analysis

This course will introduce students to a variety of datasets and teach them (hands on usage of) SAS to implement various quantitative techniques. This is an applied course that involves extensive use of data and PC-based analysis using JMP/SAS, a popular statistical software. The course will cover a number of quantitative analyses, explanatory and predictive models pertinent to marketing, such as customer segmentation, customer choice models, customer lifetime value, conjoint models, and market response models.

Course Goals
Understand how the “first principles” of marketing strategy helps firms organize the analytics opportunity and challenge in today’s data era, and use and execute data analytic techniques, and case studies to understand how to solve marketing analytics problems in a scientific and process-driven manner. Using statistical software to estimate various marketing models. Apply your learning through real database/marketing engineering cases and data.

Topics covered:
1. 4 Marketing Analytics Principles
2. JMP Orientation
3. Segmentation Concept Demo and Case
4. Targeting Concept
5. Choice Models Demo and Case
6. Customer Lifetime Value
7. Conjoint Concept Demo and Case
8. Satisfaction Analytics Concept and Case
9. Response Models Concept and Case

In MKTG 625, students undertake a several case studies focused on
• Retail site location decisions
• B2B segmentation in healthcare
• Customer acquisition in B2B, retailing and
   technology products
• Customer lifetime value in the hospitality business
• New product development in technology, and
   industrial manufacturing
• Satisfaction profit chains in the energy sector
• Market response models in the media industry

The students will learn:
1. Understanding of self and others: Behavior assessment.
2. Presence and presentation skills.
3. Business writing.
4. More on teams.
5. Emotional intelligence.
6. Leadership/fellowship.
7. Constructive criticism.
8. Peer review on capstone drafts with project coaches.
9. Industry guest speakers--applying predictive modeling.

Summer Semester (Year one) 

1. Producing predictive models
2. Work with live case studies

This course is an introduction to the general concepts and methodologies associated with Data Mining, Neural Networks, Machine Learning, and Analytics Modeling. Data Mining is the modeling and analysis of data, usually very large datasets, for decision making. Although several software packages used for Data Mining will be reviewed and compared, the primary concepts will be illustrated using SAS Enterprise Miner. Models discussed include neural networks; multiple and logistic regression; decision trees; and clustering algorithms.

Topics:
1. The data mining process
2. Introduction to SAS Enterprise Miner
3. Data collection, exploration and pre-processing
4. Linear and logistic regression in Enterprise Miner
5. High performance Enterprise Miner
6. Comparing and evaluating big data models
7. Machine Learning: Decision Trees
8. Machine Learning: Random Forests
9. Ensemble modeling
10. Introduction to text analytics
11. Document classification & sentiment analysis
12. Topic analysis
13. Machine Learning: Neural networks & Random Forests.

This course will examine the internal uses of financial and operational information in planning, controlling, decision making, and performance evaluation in the global market. A specific emphasis will be placed on learning concepts pertaining to: cost management and organizational strategy; cost behaviors; product cost flows; forecasting, cost prediction, managerial incentives and budgeting; and managerial and segment performance evaluation.

Topics discussed in ACCT 610 include
1. Product costing methods
2. Responsibility accounting
3. Operational and capital budgeting
4. Use of accounting metrics in executive
    compensation contracts
5. Firm valuation
6. Cost analysis for decision making
7. Activity-based costing
8. Balanced scorecards
9. Product and service pricing
 
Previous case studies have focused on:

1. Financial Statement Analysis
2. Product line go/no go decisions
3. Profitability measurements of individual products in a
    multi-product entity
4. Budgeting including cash budgets and pro-form
    financial statements
5. Activity-based costing
6. CRM systems and identification of profit potential of
    individual customers and distribution channels
7. Transfer pricing
8. Managerial control and incentives

The student will learn:
1. Cyber security
2. Capstone work with project coaches
3. Industry guest speakers--applying predictive modeling

Fall Semester (Year two) 

Prepare time series data for model fitting. Identify whether a time series exhibits the following properties: stationarity vs trend and/or seasonality. Fit appropriate models to time series data from a wide variety of business settings.

At the completion of the course, students will be able to:
1. Prepare time series data for model fitting
2. Identify whether a time series exhibits the
    following properties:
           a) Stationarity vs trend and/or
               seasonality
           b) Outliers and/or level shifts
3. Fit appropriate models to time series data from
    a wide variety of business settings
6. Interpret the output from models for time series data
7. Identify weaknesses in models for time series data
    and formulate ways to overcome them
8. Make valid predictions of future values and draw
    business conclusions on the basis of the fitted
    models for time series data
9. Recommend business actions on the basis of
    these conclusions

Topic list:
1. Introduction – Autocorrelation and stationarity vs trend
    and/or seasonality
2. Autoregressive (AR) models
3. Moving average (MA) models
4. Autoregressive moving average (ARMA) models
5. Autoregressive integrated moving average (ARIMA)
    models
6. Exponential smoothing
7. Regression models with autocorrelated errors
8. SAS Forecast Studio
9. Models for more than one time series
      i. Transfer function models
      ii. Multivariate time series models

Past student group projects include:
1. Based on point-of-sale grocery store sales data, forecast
   weekly ground turkey sales by U.S. region.
2. ARIMA models for theft and burglary rates
3. Seasonal ARIMA models for Monthly Economic Value of
    Mineral, Oil, and Gas Field Machinery Manufacturing
4.Predicting cattle slaughter counts
5. Seasonal ARIMA models for the number of social
    security claims received by region
6. Predicting Ready Mix Daily Sales for a large
    metropolitan area
7. Seasonal ARIMA models for average monthly energy
    bill for a large metropolitan city.
8. Predicting weekly candy sales across the US
9. Transfer function models of global mean
     temperature change as a function of
     CO2 concentration levels

This course focuses on SAS programming in the Data step, including modifying variables, formats, and loops; various procedures, including sort, means, tabulate, and freq; creating graphs; debugging; the SAS/SQL language; macros; and optimization.

Prerequisite: STAT 608
The Optimization Section primarily focuses on formulating and solving mathematical optimization using the SAS OPTMODEL procedure, from reading data to interpreting output and creating data sets. The course covers applications of linear programming, integer and mixed integer programming, and non-linear programming. Students will understand and undertake case studies. At the conclusion of the course students will be knowledgeable about optimization methodology and the use of the advanced SAS optimization capabilities.

Course topics:
1. Loops, Variable Types, Dates, Functions, Formats.
    Importing and Exporting data SAS/SQL
2. Macro language
3. Graphing Data
4. Creating HTML, PDF, RTF, and Excel output
5. Advanced data step features: BY group processing,
    indexing, arrays, multi-dimensional arrays, and more
6. optimization and PROC OPTMODEL
7. Arrays and Index Sets/LP Models
8. Formatted Output/Dual Variables
9. Control Flow, Operators, & Model Updates
10. Network Problems/ILP and MILP Models
11. Binary Variables/Separable NLP Models
12. General NLP Models/Local Search

The student will learn:
1. Industry guest speakers--applying predictive modeling.
2. capstone project work sessions with project coach.

Spring Semester (Year two) 

1. Capstone project employer and peer-presentations.
2. Graduation

 Develop a strong understanding of spatial and spatio-temporal data analysis. Learn and understand the various empirical and graphical tools for understanding spatial and spatio-temporal data that arise in business applications.

List of Topics:
1. Introduction
2. Auto correlation
3. Stationary, isotropic random fields
4. Variograms
5. Kriging
6. Estimation methods for covariance parameters
7. Spatial regression
8. Nonstationary models
9. Spatio-temporal models
10. Multivariate spatial models
Projects:
1. Modeling gasoline prices in large cities across the
    United States
2. Modeling real estate prices in King County, Washington

**Electives vary by cohort and won't be announced until fall semester of year two.

Learn how to build sophisticated financial models that analyze the impact of proposed corporate projects, investments, and other strategic decisions on shareholder value. Master the basic finance theory that underlies valuation models.

Throughout the class students will:
1. Fundamental principles underlying all financial
    models, including the concept of free cash flow,
    the time value of money and net present value.
2. The fundamentals of capital structure and
     financing policy
3. The weighted average cost of capital and adjusted
     present value methods
4. Building deterministic models for forecasting
    financial statements and free cash flows 
5. Incorporating uncertainty and simulation techniques
    into financial statement and free cash flow
    forecasting models
6. Fitting distributions for simulation of financial
    time series and other stochastic processes
7. Estimation of the cost of equity and equity capital,
    as well as the weighted average cost of capital
8. Incorporating real options into financial models
9. Applications

The ISTM 601 course provides an opportunity to build upon the introductory Python coding experience gained in the first semester (ISTM 615). Students will explore various publically available Python code libraries that are useful to data science (e.g., Pandas, NumPy, SciKit-Learn, Seaborn, etc.) and learn how to apply them in their analytics work. Other potential topics include web scraping and machine learning. Students will complete a data analysis project by writing Pyton code.

 

The student will learn:
1. Executive coaching sessions:
          a) LinkedIn profiles.
          b) Resumes for experienced professionals.
          c) Managing your professional career.
2. Final capstone presentation with graduate committee.
3. Peer and employer capstone presentations.