Deep neural networks
General data
Course ID: | 1000-2M16GSN |
Erasmus code / ISCED: |
11.3
|
Course title: | Deep neural networks |
Name in Polish: | Głębokie sieci neuronowe (wspólne z 1000-317bDNN) |
Organizational unit: | Faculty of Mathematics, Informatics, and Mechanics |
Course groups: |
(in Polish) Przedmioty obieralne na studiach drugiego stopnia na kierunku bioinformatyka Elective courses for Computer Science Specific programme courses of 2nd stage Bioinformatics |
ECTS credit allocation (and other scores): |
(not available)
|
Language: | English |
Type of course: | elective monographs |
Prerequisites (description): | Proficiency in Python programming. |
Short description: |
The goal of the course is to show use cases for deep neural networks. During the course state-of-the-art techniques, algorithms and tools will be presented. Among others two main blocks of the course will concern image classification and text processing. |
Full description: |
1.Introduction to machine learning (2 lectures): what machine learning is, supervised and unsupervised learning, regression, classification, loss function. Linear and logistic regression. Regularization, optimizing hyperparameters, judging the quality of a model. 2.Introduction to neural networks (2 lectures): introducing pytorch, how to initialize a neural network, activation functions, regularization, optimizing loss function. 3.Convolutional netural networks (4-5 lectures): image classification, benchmark datasets (MNIST, CIFAR, ImageNet), data augmentation, convlolutions, pooling, basic architectures. 4.Recurrrent neural networks (2-3 lectures): LSTM, text processing. 5.Possible futher topics depending on students' preferences (3-4 lectures). |
Bibliography: |
Online books http://neuralnetworksanddeeplearning.com/ http://www.deeplearningbook.org/ |
Learning outcomes: |
Knowledge * Knows the basics of machine learning. * Understands the learning algorithms for neural networks. * Knows basic architectures of convolutional and recurrent neural networks. Abilities * Is able to use a chosen modern library of machine learning procedures. * Can implement image classification algorithms based on convolutional neural networks. * Can implement text processing algorithms based on recurrent neural networks. * Can use English at the proficiency level B2+ of Common European Framework of Reference for Languages, with particular emphasis on the computer science terminology Competences * Is ready to critically evaluate acquired knowledge and information. * Is ready to recognize the significance of knowledge in solving cognitive and practical problems and the importance of consulting experts when difficulties arise in finding a self-devised solution. * Is ready to think and act in an entrepreneurial way. |
Assessment methods and assessment criteria: |
Final grade is based on the sum of scores for the project, homework assignments and final exam. Homework assignment consists of a few programs to be implemented. |
Copyright by University of Warsaw.