• A
  • A
  • A
  • АБB
  • АБB
  • АБB
  • А
  • А
  • А
  • А
  • А
Обычная версия сайта
Статья
Применение технологии Process Mining в управлении цепями поставок

Воронова А. П., Заходякин Г. В.

Логистика и управление цепями поставок. 2020. № 6(101). С. 26-36.

Глава в книге
Conceptual Framework of Agent-based Model of Relational Conflicts in Russian Retail

Morozova Y. A.

In bk.: Proceedings of Analytics for Management and Economics Conference AMEC 2019. St. Petersburg: 2019. P. 367-373.

Predictive Analytics for Logistics and Supply Chain Management

2019/2020
Учебный год
ENG
Обучение ведется на английском языке
3
Кредиты
Статус:
Курс по выбору
Когда читается:
2-й курс, 1 модуль

Преподаватель

Course Syllabus

Abstract

This is an elective course, offered during the 2nd year of the MSc program “Strategic Logistics Management”. The goal of this course is to provide an overview of classical and modern methods for solving business problems that can be addressed with predictive analytics and to extend the practical skills of predictive modeling. The course builds upon knowledge gained in the research seminar and extends it towards modern modeling methods such as ensemble models and deep learning. The topics of model selection, feature selection and feature engineering are covered in greater detail. All classes run in a computer lab. A class includes an overview of theoretical principles, a guide to implementation of analysis methods in software and a hands-on example for solving a practical problem using software tools. This is followed by a discussion of practical application of predictive methods in business analytics. In order to pass the course, all participants are required to give a short presentation on application of predictive analytics tools and to do an individual or group project for application of predictive analytics tools to solve a real-world problem of their choice (students can join a data science competition, for example). Pre-requisites The participants should have basic knowledge of data processing tools and technologies, as well as understanding of modeling methods such as regression. Research Seminar “Methods and tools of Data Analysis” offered in the 3rd year of the BSc program “Logistics and Supply Chain Management” would be a good background. For students who didn’t have the opportunity to take that course several competences are required: • working knowledge of descriptive statistics and exploratory data analysis (graphs and descriptive statistics); • working knowledge of analytical statistics (statistical hypotheses testing, correlation and regression); • familiarity with programming languages for statistical data analysis (R/tidyverse is preferable); • basic understanding of operations with relational data, such as filtering, sorting, grouping and joining data tables.
Learning Objectives

Learning Objectives

  • To explain the application of predictive models in a business analytics context
  • To describe the steps required to develop an analytical solution
  • To define predictive analytics task types
  • To elaborate on methods for solving regression and classification type problems
  • To explain methods and approaches for assessing and improving the quality of predictive models
  • To provide tutorial on implementing data analysis steps using R
  • To discuss business cases and opportunities of predictive modeling
Expected Learning Outcomes

Expected Learning Outcomes

  • to define the predictive analytics concept
  • to name use cases of predictive modeling in Logistics and Supply Chain Management
  • to list the steps in implementation of a predictive analytics solution
  • to develop a prototype analytical solution to a business problem using R and to present its outcomes to decision makers
  • to definde the concept of statistical/machine learning
  • to explain the tradeoff between model bias and variance
  • to identify a predictive task type based on a problem formulation
  • to construct linear predictive models using R
  • to interpret model accuracy metrics for regression models
  • to prepare data for a predictive model
  • to implement linear classifiers using R
  • to interpret performance metrics for classification models
  • to apply resampling strategies for model evaluation
  • to justify an appropriate model type and specification using model benchmarking techniques
  • to apply variable selection procedures for model specification
  • to construct regularized linear models using R
  • to construct tree-based regresssion and classification models using R
  • to apply the MLR framework for constructing, benchmarking and tuning predictive models
  • to construct ensemble-based models using R
  • To implement deep learning models using Keras and R
Course Contents

Course Contents

  • Topic 1. Predictive analytics defined. Use-cases. Lifecycle of an analytical solution.
    Evolution of business analytics methods and technology. Predictive analytics defined. Predictive task types. Use-cases and opportunities for predictive analytics in Logistics and SCM. Lifecycle of an analytical solution. The CRISP-DM process.
  • Topic 2. The concept of statistical/machine learning
    Statistical and Machine Learning. Types of learning. Measuring model accuracy. Tradeoffs between model flexibility, interpretability and accuracy.
  • Topic 3. Regression using linear models
    Regression using linear models. Model building, quality assessment, inference. Modeling non-linear relationships and interactions. Predictor importance assessment for regression problems. Applications of regression methods for logistics and supply chain management.
  • Topic 4. Classification using logistic regression
    The classification task. Classification using logistic regression. Assessing classification quality. Predictor importance assessment.
  • Topic 5. Resampling for model performance assessment
    Estimating the model generalization error using resampling. Resampling strategies.
  • Topic 6. Variable selection and model regularization.
    Variable selection and model regularization. Penalized loss functions. Variable selection strategies. Lasso, Ridge and Elastic Net regression.
  • Topic 7. Tree-based classifiers
    Tree-based classifiers. Classification with highly imbalanced classes. Comparing classifiers. Cost-based classification. Selecting the threshold for soft-classifiers. Comparing performance of classifiers. Applications of classification methods for logistics and supply chain management.
  • Topic 8. Modeling frameworks
    The MLR framework for predictive model building
  • Topic 9. Ensemble models
    Meta-learning and ensemble methods: bagging, Random Forest, Boosting.
  • Topic 10. Deep Learning
    Introduction to Deep Learning. The evolution of artificial neural networks. A Multilayer Perceptron. Learning algorithms for neural networks. Network Architectures for Deep Learning. Implementing neural networks using Keras. Using neural networks for regression and classification. Application of Deep Learning for computer vision and time series analysis.
Assessment Elements

Assessment Elements

  • non-blocking Participation
    The grade is computed by dividing the sum of points for participation obtained by the student divided by the total points for all assignments posted and multiplied by 10. The result is posted into the gradebook without rounding.
  • non-blocking Presentation
  • blocking Project Defense
Interim Assessment

Interim Assessment

  • Interim assessment (1 module)
    0.25 * Participation + 0.25 * Presentation + 0.5 * Project Defense
Bibliography

Bibliography

Recommended Core Bibliography

  • Gareth James, Daniela Witten, Trevor Hastie, Rob Tibshirani, & Maintainer Trevor Hastie. (2013). Type Package Title Data for An Introduction to Statistical Learning with Applications in R Version 1.0. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsbas&AN=edsbas.28D80286
  • Hastie, T., Tibshirani, R., & Friedman, J. H. (2009). The Elements of Statistical Learning : Data Mining, Inference, and Prediction (Vol. Second edition, corrected 7th printing). New York: Springer. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=277008
  • James, G. et al. An introduction to statistical learning. – Springer, 2013. – 426 pp.
  • Lantz, B. (2019). Machine Learning with R : Expert Techniques for Predictive Modeling, 3rd Edition (Vol. Third edition). Birmingham, UK: Packt Publishing. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=2106304
  • Larose, D. T., & Larose, C. D. (2015). Data Mining and Predictive Analytics. Hoboken, New Jersey: Wiley. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=958471
  • Taweh Beysolow II. (2017). Introduction to Deep Learning Using R. Springer. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsrep&AN=edsrep.b.spr.sprbok.978.1.4842.2734.3

Recommended Additional Bibliography

  • Hastie, T., Tibshirani, R., Friedman, J. The elements of statistical learning: Data Mining, Inference, and Prediction. – Springer, 2009. – 745 pp.
  • Larose, D. T., Larose, C. D. Discovering knowledge in data: an introduction to data mining. – John Wiley & Sons, 2014. – 336 pp.