About the Acadmic Track

About the Acadmic Track

Master of Statistics

The Master in Statistics focuses on data analysis, methodological problem-solving in a range of disciplines, and various types of statistics software, or mathematical statistics. The academic track is open to experts wishing to improve their theoretical skills in statistical research, but also to non-experts wishing to develop their data analysis skills in an area of specialization acquired at the Bachelor level.
CAREER OPPORTUNITIES
The Master of Science in Statistics opens doors to opportunities in various industries: in the food industry, international organizations, in (cantonal and federal) statistics offices, in financial companies, biostatistics (in the pharmaceutical industry), and in many other areas. Finally, students who have completed the master can continue with a doctorate.

Study System

Study Duration

● The student studies eight courses, distributed as follows:
√ Six compulsory courses.
√ Two elective courses from among the courses offered by the college for master’s students.
● The study is conducted through research seminars in each course. The research seminar is based on multiple references and is in accordance with the methodology and standards of scientific research.
● The study of each of the eight courses takes four credit hours for a minimum of four weeks and may exceed that according to each student’s abilities. After that, the student’s competency and knowledge test is held in the course he completed, then he begins another course in the same manner, and so on.
● The student is assigned a hypothetical course that the college chooses from among the courses that the student has studied at the undergraduate level, and this is considered a practical training for the student to be evaluated with ten credit hours. The student must divide this course from twelve to fourteen brief lectures. The student presents each lecture in the form of a written summary of its topic in Word or PDF format, accompanied by a video recording of it in the student’s voice using the Power Point program, with a duration of not less than ten minutes and not more than About twenty minutes.
● Study courses in the first year, the student has the right to extend the study for a period not exceeding a second year.

The requirements for obtaining a master’s degree in various disciplines are sixty credit hours according to the study plan approved by the University Council, and these requirements are distributed according to the following programs:
1- Research courses of thirty-two credit hours.
2- A scientific thesis with eighteen credit hours.
3- A practical training of ten credit hours.

Requirements for registering a thesis topic for a master’s degree
● The student must pass the stipulated courses with a score of at least 70%.
● The student obtained a TOEFL certificate of at least 450, or its equivalent, or obtained an equivalent certificate in the French language, with the exception of those who obtained a first university degree in one of the two languages, or in one of the two languages.
● The student submits a request to the university administration to register a master’s thesis with a proposed topic in one of the subspecialty tracks.
● If the initial approval of the subject title is achieved, the college council will designate a supervisor to guide and follow up the student in preparing the plan.
● The research plan includes the importance of the topic and a critical presentation of previous studies in it, specifically the research problem, then defining the study’s curriculum and its main hypotheses or questions that you want to answer, and the division of the study and its sources.
● The student presents his proposed plan in a scientific seminar, discussing the plan as a topic and an approach.
● The student adjusts his plan based on the professors’ observations in the seminar if he is asked to amend.
● After the seminar, the plan is presented to the college council to take its decision regarding the registration of the subject.
● In the event of approval, the College Council’s decision is presented to the University Council to approve the registration, and the registration date is calculated from the date of the University Council’s approval.

Jury discussion and award of degree
● The minimum for preparing a master’s thesis is nine months, starting from the date of approval by the University Council to register the subject, and the maximum is two years, which can be extended for a third exceptional year upon the recommendation of the supervisor and the approval of the College Council, provided that the total period of student enrollment in the degree does not exceed four years.
● The supervisor submits a semi-annual report that includes what has been accomplished and what is required in the remaining period.
● After the student completes the thesis and the supervisor reviews it, the supervisor submits to the university administration a report stating its validity for discussion, including an evaluation of the student’s performance during the preparation period of the thesis of 140 degrees, with a full copy of the thesis signed by him, and a letter with the names of the discussion and judgment committee proposed by the professors of the specialty, for presentation to the college Council.
● At least fifteen days must pass before the student’s discussion from the date of the approval of the discussion committee by the college.
● The validity period of the committee formed to discuss the thesis is six months, which may be renewed for a similar period based on a report from the supervisor and the approval of the College Council.
● Each member of the committee writes a detailed scientific report on the validity of the thesis for discussion, and the thesis is evaluated out of 420 degrees, and the average of the three degrees is taken.
The student may not be discussed unless he/she gets at least 70% of the supervisor’s evaluation of his performance and the committee members’ evaluation of the message in the individual reports.
● A group report is submitted after the discussion, signed by all members of the committee, in which an evaluation of the thesis discussion is given on a scale of 140 degrees.

The thesis is passed after public discussion with one of the ratings shown in the following table:

Percentage of gradespointsappreciation symbolAppreciation
ArabicEinglish
95 to 100%4A+A+Prominent
90 to less than 95%7 , 3aA
85 to less than 90%3 , 3b+BVery well
80 to less than 85%3BB
75 to less than 80%7 , 2c+C+Good
70 to less than 75%3 , 2cC

After the college approves the student’s results, the master’s degree is awarded at a rate calculated from the average total of the courses and thesis grades.
After obtaining the approval of the University Council to grant a master’s degree to the student, he is entitled to obtain insured certificates, authenticated by the university, stating that he obtained that degree, in order to present them to the various authorities.

Study Duration

The duration of study to obtain a master’s degree in Political and economics is two years as a minimum, and six years as a maximum.
In the first year, the student studies at least eight subjects, and the study is through research seminars for each course. The research seminar is based on multiple references and is in accordance with the scientific research methodology and standards.
In the second year, the student attends a general seminar for the topic of the thesis, which he will prepare and submit for discussion
The general seminar is discussed by the scientific committee at the university, and the title of the thesis is approved
The student works to complete his thesis under the supervision of the supervisor decided by the Presidency of the University based on the proposal of the Dean of the Faculty
The student completes his scientific thesis and submits for discussion before the committee formed by the Presidency of the University in a public session and completes the conditions for a master’s degree
Courses of study in the first year The student has the right to extend the study in it for a period not exceeding a second year
The thesis prepared by the student during a period of time not less than 9 months and not exceeding two years

Conditions for success and graduation

1) The student is considered to have passed any of the program’s courses if he achieves a final score of no less than 65%. He is also considered successful in the master’s project if he obtains a mark (granted by the judging committee) not less than 75%.
After the student presents the results of his project before the committee, and discusses its technical content.
2) The student is not entitled to submit to discuss his thesis until a scientific research is published in an approved refereed journal.
3) The student obtains a master’s degree certificate after he has fulfilled all the scientific requirements for this degree.

Acadmic Track Structure

Acadmic Track Structure
8 courses = 32 credit hours,
 practical training = 10 credit hours
 Master's thesis = 18 credit hours
Courses
Practical Training
Master's Thesis

Curriculum

I. Compulsory Courses

Scientific Research Methodology

Course name: Scientific Research Methodology

Course code: MECO101

Credit hours: 4.00

The course aims to train students to prepare research in economics science, by identifying the most important steps of the research process such as the research problem, scientific hypotheses, concepts and variables, data collection methods, data analysis tools, reaching results and generalizations, in addition to introducing the most important approaches used in field of economics science.
The curriculum’s inputs include: the form and type of knowledge, learners’ characteristics, needs, tendencies and interests, the society’s philosophy, values, hopes and aspirations. The outputs of the curriculum are: knowledge, skills, and attitudes.

Applied Bayesian Statistics

Course name: Applied Bayesian Statistics

Course code: MECO102

Credit hours: 4.00

Objective
At the end of the course, a successful student is expected to be able to:
– understand the underpinning ideas of Bayesian statistics
– model a problem using the Bayesian paradigm
– simulate with the software R random variable/vector using various simulation techniques
– compute summaries of posterior distributions using the software R
– interpret the results of a Bayesian modelling
– discuss the weaknesses and advantages of a given Bayesian model

Description
The course aims at giving an introduction to Bayesian statistics, with focus on applications (using the software R) rather than on theory, though some theory might be introduced to justify the methods seen. The plan is to cover the following topics (subject to minor changes):
– Bayes Theorem and the Bayesian Paradigm
– Priors, conjugate priors, posterior
– Summaries of distributions via simulations
– Simulation techniques (inversion, importance sampling, MCMC methods)
– Bayesian (generalized) linear models, Bayesian mixture models
– Bayesian model choice, uncertainty, and model averaging
– Naive Bayes, Bayesian networks and Bayesian Decision Analysis
– Gaussian Processes
– variational inference
Some advanced topic might be presented in the last lecture, such as the Doob and Bernstein–von Mises theorems, and constrained optimisation generalisations of Bayes theorem, but these will not be examinable.

Modern Flexible Regression

Course name: Modern Flexible Regression

Course code: MESA103

Credit hours: 4.00

Objective
Provide a set of tools for flexible regression analysis of the relationship between measurements made on (groups of) subjects or objects for large classes of response distributions (e.g. nominal, ordinal and continuous data) and settings. At the end of this course, a student should be able to decide which technique is relevant in which situation and to successfully apply it, including to new situations.

Description
This course will introduce approaches to regression modelling for a large variety of continuous and discrete response variables. The cases of independent as well as correlated responses will be addressed.
The first part of the course will be devoted to univariate nonparametric regression, mostly via splines, even though we will briefly mention other alternatives.
The second part of the course will deal with generalized models for independent responses, in their linear parametric form (GLM) and in their nonparametric additive form (GAM). We will consider several settings within the exponential family of distributions (e.g. binomial, Poisson, Gamma,’) and beyond (negative binomial, beta,…).
Finally, the third part of the course will consider extensions of these models. To model non independent responses, we will look at GEE, GLMM and GAMM models. To extend the flexibility of the structural relationships, GAMLSS will come at rescue.
Many real data applications will illustrate the use of the models introduced and will be analyzed during the exercises sessions (labs), which are part of the course.
The R software will be used throughout.

Linear Models for Dependent Data

Course name: Linear Models for Dependent Data

Course code: MESA104

Credit hours: 4.00

Objective
Upon success, the student will be able to fit mixed linear models to data which design applies in the context of random effects, interpret the model outputs and perform statistical inference from it, build appropriate experimental designs for experiments to which these models can apply, and discuss mathematical and statistical issues associated with these models.

Description
The content of the course is:
– Design: data structure, factors, nested design.
– Exploratory data graphical tools.
– Model: Fixed and random effects, marginal/primal effects and interactions
– Estimation: maximum likelihood and REML
– Inference: likelihood ratio test, variance decomposition, ANOVA and bootstrap.
– Prediction: at the population level, at the individual level, and model diagnostics.
The theory is illustrated on several data sets covering repeated measure ANOVA, growth curves and variance functions. Exercises and practical cases are done using the computer program R.

Multivariate Analysis

Course name: Multivariate Analysis

Course code: MESA105

Credit hours: 4.00

Objective
Upon success the student will be able to:
– Choose appropriate methods among the ones enumerated in the content.
– Perform the corresponding statistical analysis using the computer program R.
– Discuss advantages and drawbacks of each method.
– Discuss mathematical issues associated to each method.

Description
The course covers the topics :
– Multivariate probability distributions
– Multivariate testing and confidence sets
– Principal Component Analysis
– Cluster analysis
– Kernel methods
– Graphical models
– Missing data

The Statistical Analysis of Time Series

Course name: The Statistical Analysis of Time Series

Course code: MESA106

Credit hours: 4.00

Objective
Students finishing this course successfully are expected to:(1) be familiar with the theory required for the statistical analysis of time series;(2) understand and apply various inferential techniques, e.g. maximum likelihood estimation, and spectral analysis;(3) critically evaluate the outcomes of the applied inferential techniques;(4) be familiar with the statistical software for the analysis of real data; (5) be able to present the results of a statistical analysis.

Description
Time series analysis refers to phenomena in which observations are collected over time and there are correlations among successive observations. Applications cover virtually all scientific areas, including economic, finance, medicine, and climatology. This course provides the Students with the most important concepts and methodologies for dealing with time series analysis. The course starts from the fundamental definitions(e.g., stationarity and ergodicity) and it covers topics in both time- and frequency domain. In time-domain, both univariate (e.g. autoregressive processes) and multivariate time series models are discussed. In frequency-domain, spectral methods are applied to study long-memory (e.g. autoregressive fractional integrated) processes and to analyze large data sets (dimensionality reduction). Finally, an overview about Hilbert spaces approach to time series concludes the course. The theoretical aspects are illustrated by the practical analysis of data sets. A brief review of some (high-school) math skills (e.g. trigonometry, complex numbers, matrix calculus) is done in the exercise sections.

II. Elective Courses

 

Advanced Statistical Inference

Course name: Advanced Statistical Inference

Course code: MESA151

Credit hours: 4.00

Objective
Learning Outcomes:
By the end of the course, the student must have the technical skills to understand complex inferential problems. In particular, they must be able to:
‘ Formulate the various elements of a statistical problem in a rigorous way.
‘ Assess the performance of estimators.
‘ Construct efficient statistical procedures for estimation.
‘ Understand advanced statistical techniques to extract efficiently the information
content of complex datasets.
‘ Specify suitable tools for data analysis in several scientific areas.
‘ Identify new challenges related to data analysis.

Description
Statistical inference is presented by covering the fundamental theory and method at an advanced level. The course has two parts. In the first part (the main one), we start from a review of some fundamental results in probability and measure theory.
Then, we move to the key notions related to data reduction techniques (sufficiency, completeness and ancillarity) and to the general theory of estimation (unbiasedness and optimality). Asymptotic (large sample) theory, a crucial part of statistical inference, is studied throughout the course, rather than in a separate/specific lecture. In the second part, we give an overview of some recent developments in the inference for dependent data (spatial processes and network data).

Data-Driven Impact Evaluation

Course name: Data-Driven Impact Evaluation

Course code: MECO152

Credit hours: 4.00

The objective of this course is to introduce the basic concepts of causal inference.
The emphasis is on quantitative methods for impact evaluation and treatment effect estimation.

Description
– Potential Outcomes, Causality and Experiments
– Matching and Regression
– Propensity score weighting and regression
– Differences in Differences
– Using Instrumental Variables
– Discontinuity Regression
– Non- and semiparametric regression techniques

Fundamental and Advanced Sampling Techniques

Course name: Fundamental and Advanced Sampling Techniques

Course code: MESA253

Credit hours: 4.00

Objective
This course gives an overview of sampling theory, on which is based a large part of official statistics produced by governments or international organizations such as Eurostat, OECD, etc. Official statistics play a major role in our societies as they provide important information to policy and decision makers. At the end of the course, students should be able to 1) understand how official statistics are produced, 2) identify major challenges occurring in sample surveys, 3) find proper solutions.

Description
We first introduce some basic concepts of sampling theory, as well as some usual sampling designs. Then we highlight that producing results based on data that stem from a random sample survey involves several steps of treatment and analysis. For example, difficulties occurring in practice during data collection, such as total or partial non-response, require adapted treatments. Moreover, we can nowadays take advantage of a wide range of auxiliary information thanks to an increasing availability of administrative data or other data sources. These additional information sources may help to improve the precision of survey estimators, thanks to a technique called ‘calibration’ which will be discussed during the course. We will also present variance estimators adapted to these procedures. Coordination of survey samples is an important topic in statistical agencies. We will present the most common techniques of sample coordination currently in use and illustrate through examples how to implement them in practice. The course is delivered by methodology experts of the Statistical Methods Unit of the Swiss Federal Statistical Office in the form of lectures and exercises inspired by their own experience.

Experimental Design : Theory and Practice

Course name: Experimental Design : Theory and Practice

Course code: MESA254

Credit hours: 4.00

Objective
This course presents the key differences between observational and experimental approaches when aiming for causation more than association.
The main objective is to introduce most efficient experimental designs that, in parallel with simple linear models, allow to get conclusive results out of very small datasets. Multiple examples from food science and nutrition & health business are used to show the strength of this approach vs. using complex algorithms on big observational datasets.

Description
Successful Innovation & Renovation of consumer goods strongly relies on sound models relating many dependent variables with many independent variables.
The importance of linear models (t-test, ANOVA, mixed models, regression) and inference techniques (both parametric and non-parametric) is exemplified through the multivariate modeling of consumer acceptance, health benefits and sensory properties of food products as function of their recipes (i.e. raw materials, formulation, process).
Design techniques to properly test the performance of many products (e.g. parallel, crossover, Williams Latin squares designs) are then discussed in the frame of consumer and sensory sciences as well as on health outcomes.
Design techniques that allow to properly model the causal effects of recipe parameters on product characteristics are then discussed in detail. This part starts with full factorial designs with ‘2 levels, followed by fractional factorial designs, asymmetrical fractional factorial designs, Plackett-Burman designs, supersaturated designs and orthogonal arrays.
Finally, response surface modeling, optimality criteria (e.g. D-, A-, space-filling) and Taguchi approaches are introduced before addressing mixture designs and the combination of mixture and fractional factors.

Financial Econometrics

Course name: Financial Econometrics

Course code: MECO255

Credit hours: 4.00

Objective
– Develop the necessary econometrics tools for financial applications
– Applications of parametric and non-parametric statistics
– Implementation of econometrics tools through MATLAB

Description
The course will develop the econometrics tools and models in the context of financial markets in static and dynamic environment.
The first part will covers static models and tools such as CAPM, APT, Value at Risk, Expected Shortfall and scoring procedure, while the second part will focus on dynamic analysis build around ARMA models, GARCH models and Stochastic volatility models. Thorough the course, statistical tools such as linear regression, asymptotic theory, non-parametric estimation and bootstrap procedure will be discussed in detail. Illustration will be discusses in class. The seminar will implement those methods within the framework of MATLAB

Model Selection in High Dimensions

Course name: Model Selection in High Dimensions

Course code: MESA256

Credit hours: 4.00

Objective
The learning objectives are for the student to 1) acquire the necessary fundamental notions for model selection such as out-of-sample validity, prediction error, over- and under-fitting, etc. 2) to study regularized regression methods such as the (adaptive) lasso, MCP, SCAD, elastic net, 3) practice algorithms/methods such as cross-validation, CART, LARS, OMP, SIS, stepwise/streamwise/stagewise search, 4) to put into practice the methods through the analysis of a (big) dataset (project)

Description
Model selection in high dimensions is an active subject of research, ranging from machine learning and/or artificial intelligence algorithms, to statistical inference, and sometimes a mix of the two. We focus on the frequentist approach to model selection in view of presenting methods that have the necessary properties for out-of-sample (or population) validity, within an as large as possible theoretical framework that enables the measurement of different aspects of the validity concept. We therefore anchor the content into an inferential statistics approach, essentially for causal models.

More specifically, the focus of model selection in high dimensions is presented into two main headings, one on statistical methods or criteria for measuring the statistical validity, and the other one on fast algorithms in high dimensional settings, both in the number of observation and in the number of inputs, that avoid the simultaneous comparison of all possible models.

The models are mainly the linear regression and the logistic regression (for classification) and the domains of applications range from economics to medical sciences.

An important part of the class is devoted to the practice of model selection in high dimensions methods, using R packages, that includes a semester project that is used for the final evaluation.

b17
Degree: Master's Degree
Track code: MA110PE
Study method: Distance Learning
Credit hour: 60
How long it takes: 
Full time: 2 years
Part time: 4 years
Limit time: 6 years
Apply Now