Computer Science

The course introduces basic theory and algorithms of machine learning. Topics include: supervised learning setting; unsupervised learning setting; concentration of measure inequalities; analysis of generalization in classification; algorithms; assumptions behind the algorithms taught in the course, their implications, and common pitfalls; and correlation versus causality. The course assumes solid math and programming skills, including knowledge of linear algebra, calculus, probability theory, discrete mathematics, and programming. 

The Individual Research Training Senior (IRT Senior) Course is an advanced course of the Individual Research Training A (IRT A) course in the Tohoku University Junior Year Program in English (JYPE) in the fall semester. Though short-term international exchange students are not degree candidates at Tohoku University, a similar experience is offered by special arrangement. Students are required to submit: an abstract concerning the results of their IRT Senior project, a paper (A4, 20-30 pages) on their research at the end of the exchange term, and an oral presentation on the results of their IRT Senior project near the end of the term.

After completing this module, the participants have gained practical experience in designing, implementing, and testing of applications for the internet of services within a small team and therefore gained significant knowledge and insights within the areas of mobile devices, communication and services, location-based services, cloud computing and digital communities. Furthermore, the participants have gained important knowledge of how to organize and realize IT projects, including controlling, reporting, planning, and communicating with external partners. 
 

The following topics are covered in this course: computer arithmetic, number formats (place value systems, fixed- and floating-point numbers); basics of digital design ((combinatorial logic, gates, truth tables, storage elements, finite state machines); basic technologies and components of a (secure) computer architecture; assembly programming (MIPS): assembly language, control flow, addressing; structure and operation of a multi-cycle data path (MIPS), structure and operation of a multi-cycle implementation; measuring and evaluating performance (SPEC benchmarks, Amdahl's law); structure and operation of a simple Von Neumann model; introduction to pipelining: concepts, hazards, forwarding, solutions; memory hierarchy, caches, virtual memory; input/output techniques (addressing, synchronization, direct memory access).

This course is part of the Laurea Magistrale degree program and is intended for advanced level students. Enrolment is by permission of the instructor. This course introduces the main concepts of Python and its use in economic and econometric analyses. In particular, the course focuses on: 1) data types: definitions and use; 2) pandas; 3) basic programming structures (loops, if,...); 4) a primer on classes; and 5) applications to economics and econometrics.

This course discusses machine learning and its uses in business decision-making. Topics include: data extraction and exploration; basic models for classification and regression; training, hyper-parameter tuning, model evaluation, pre-processing; feature selection and generation; advanced models for classification and regression; unsupervised learning.

This is an advanced-level Data Science course, focusing on deep learning, which has witnessed great success over the past decade. Two of the most successful fields of deep learning are image processing and natural language processing. 

Some of the most successful applications of deep learning in image processing include object detection, image segmentation, and image classification. In natural language processing, deep learning has been used to develop applications such as machine translation, text classification, automatic summarization and question answering.  

The course begins with an overview of deep learning, and a review class for Python and the PyTorch library respectively. Then, the course studies linear algebra and calculus from numerical perspectives. The course also reviews the basics of statistics and information theory for deep learning and the basics of machine learning, including topics like overfitting, supervised and unsupervised learning, and stochastic gradient descent.  

The course introduces neural network models using the familiar linear and softmax regression, as well as the concept of multilayer perceptrons and the essential technique of backward propagation.  The course also studies various ways to regularize deep neural networks, such as putting norm penalties or allowing dropout, and how to do optimization for training these regularized deep neural networks. The latter half of the course focuses on convolutional neural networks for image processing and recurrent and recursive neural networks for natural language processing. Last, the recent important topic of fine-tuning a pre-trained large language model will also be covered. 

This course focuses on the practical aspects of the automated processing of human languages. It develops knowledge of useful and logical aspects, as well as useful prototypes of the same nature. The course introduces the basics of the programming language Python.

This course introduces students to modern programming techniques using the Java programming language as an example. The use of object-oriented concepts enables students to quickly work on complex tasks independently. In the practical exercises, students also learn how to use a development environment and a version management system (git) while programming. The programming language used is Java. -Java basics: * Data types, variables, operators, static methods / functions - Object orientation: * Classes and objects * Polymorphism with interfaces * Generics * Implementation inheritance - Java Collections - Error handling - Input / Output - GUI if necessary.