ISOM 2500

Business Statistics
Spring , 2018
Department of Information Systems,
Business Statistics and Operation Management
(Preliminary Draft)

Instructor: Prof. Xuhu Wan, LSK Building, Room 4072, Ext.7731, imwan@ust.hk.

Tutor: Fung Lok, imhflok@ust.hk

Class meets:

• L3: 9:00- 10:20 am , Monday and Wednesday at 2464

• L4: 10:30- 11:50 am , Monday and Wednesday at 2464

Course Description

Main Contents: This course is about how to reach a conclusion from uncertainty. We not only discuss
theoretical concepts and methods of statistics, also focus on processing real data, building regression
models with python, finally validating and testing models with new data.
Theoretically, we will discuss variables, relation between variables, conditional expectation, estima-
tion, testing of significance, regression. More than that, industry-oriented examples are given to help
students learn the problem-solving skills in real situation. For examples, we will use python to
• download stock data and preprocess data
• analyze the mean and variance of a portfolio of assets
• select the important variables for prediction model
• use regression-model based signal to trade
We will use python throughout this course, assignment and exams as well.

Objectives: •
• Measure relation between two variables and select important variables
• Apply python in data analytics
• Use and evaluate regression models
• Apply regession models in different scenarios

Course Materials

A. Lecture note ” Introduction to Statistical Analysis ”

B. All topics and assigned projects require the installation of python package ”Anaconda”.

C. A course website (http://canvas.ust.hk ) is maintained which contains lectures notes, assignments, and
links of videos and data.

Evaluation

Your overall grade will be based on the following:

A. Four assignments(50% ): Students are allowed to work individually or in group (group size ≤ 3) on the
problem sets. You are required to show your work clearly and NO credit will be given if it is not met.
All assignments are required to submit online in the format of jupyter notebooks.

B. In-class Midterm(15%): The midterm must be done independently. You can refer to notes, books and
internet, but are forbidden to communicate with other humans or AIs.
C. Final(35%): The final must be done independently. You can refer to notes, books and internet, but are
forbidden to communicate with other humans or AIs.

D. Standards for Assessment:

A+, A, A− Excellent Performance

B+, B Good Performance

B−, C+, C Marginal Performance

D Unsatisfactory Performance

Course Organization

• Part I: What Data Scientists Do

– Play with stock data using DataFrame
∗ download from google finance
∗ import from data file
– Let us make money
∗ Moving average
∗ A silly trading strategy
– Making money is not to gamble
∗ Can we predict future
∗ Build our first prediction model
∗ Evaluating your model is more important
• Part II: Variables: Predictors and Response of Prediction Models
– Data and Variables, Sample and Population
∗ Population and sample
∗ Variables and data
∗ Types of variables
∗ Distribution and empirical frequency
– Most Dry but Most Important Part
∗ Independence and conditional probability
∗ Expectation and conditional expectation
∗ Estimation of mean of log return
∗ Statistical significance and testing
– Identify Useful Predictors for Response
∗ Correlation coefficient
∗ Cramer’s V
∗ Variable selection
• Part III: Let Us Build Regression Models
– How to find best linear model
∗ Definition of best fit
∗ Least square estimation
∗ Evaluation of models
– Let Us Learn Some Theories of Regression Models
∗ Four Assumptions and why
∗ Statistical inference if four assumptions hold
∗ Whatif assumptions do not hold

2
 – Trade with multiple stock indices
– Evaluation of performance of prediction-based strategy
• Part IV: What is next

Grievance Procedure

If you disagree with grades that have been assigned to your work, you have the possibility to meet
instructors within one week after the grades have been published on the course website. Be specific about
what it is that you don’t agree with.

Academic Integrity

Academic dishonesty includes, but is not limited to, cheating, plagiarizing, fabricating of information fa-
cilitating acts of academic dishonesty by others, having unauthorized possession of examinations, submitting
work of other groups, or tampering with the academic work of other groups. All exam answers must be your
own, and you must not provide any assistance to other students during exams.