**Are you an EPFL student looking for a semester project?**

Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.

Lecture# Convolutional Neural Networks: Fundamentals

Description

This lecture covers the fundamentals of Convolutional Neural Networks (CNNs), including data statistics, training optimization, and the structure of CNN layers. Topics include neural network optimization, convolution definitions, gradient descent variants, and the importance of learning rates. The lecture also delves into the concept of pooling layers, feature engineering, and the potential pitfalls of relying solely on summary statistics.

Official source

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

In course

ME-390: Foundations of artificial intelligence

This course provides the students with 1) a set of theoretical concepts to understand the machine learning approach; and 2) a subset of the tools to use this approach for problems arising in mechanica

Instructor

Related concepts (247)

Convolutional neural network

Convolutional neural network (CNN) is a regularized type of feed-forward neural network that learns feature engineering by itself via filters (or kernel) optimization. Vanishing gradients and exploding gradients, seen during backpropagation in earlier neural networks, are prevented by using regularized weights over fewer connections. For example, for each neuron in the fully-connected layer 10,000 weights would be required for processing an image sized 100 × 100 pixels.

Convolution

In mathematics (in particular, functional analysis), convolution is a mathematical operation on two functions (f and g) that produces a third function () that expresses how the shape of one is modified by the other. The term convolution refers to both the result function and to the process of computing it. It is defined as the integral of the product of the two functions after one is reflected about the y-axis and shifted. The choice of which function is reflected and shifted before the integral does not change the integral result (see commutativity).

Low-pass filter

A low-pass filter is a filter that passes signals with a frequency lower than a selected cutoff frequency and attenuates signals with frequencies higher than the cutoff frequency. The exact frequency response of the filter depends on the filter design. The filter is sometimes called a high-cut filter, or treble-cut filter in audio applications. A low-pass filter is the complement of a high-pass filter. In optics, high-pass and low-pass may have different meanings, depending on whether referring to the frequency or wavelength of light, since these variables are inversely related.

Electronic filter

Electronic filters are a type of signal processing filter in the form of electrical circuits. This article covers those filters consisting of lumped electronic components, as opposed to distributed-element filters. That is, using components and interconnections that, in analysis, can be considered to exist at a single point. These components can be in discrete packages or part of an integrated circuit. Electronic filters remove unwanted frequency components from the applied signal, enhance wanted ones, or both.

Chebyshev filter

Chebyshev filters are analog or digital filters that have a steeper roll-off than Butterworth filters, and have either passband ripple (type I) or stopband ripple (type II). Chebyshev filters have the property that they minimize the error between the idealized and the actual filter characteristic over the operating frequency range of the filter, but they achieve this with ripples in the passband. This type of filter is named after Pafnuty Chebyshev because its mathematical characteristics are derived from Chebyshev polynomials.

Related lectures (829)

Neural Networks: Multilayer LearningPHYS-467: Machine learning for physicists

Covers the fundamentals of multilayer neural networks and deep learning, including back-propagation and network architectures like LeNet, AlexNet, and VGG-16.

Deep Learning FundamentalsME-390: Foundations of artificial intelligence

Introduces deep learning, from logistic regression to neural networks, emphasizing the need for handling non-linearly separable data.

Neural Networks: Training and ActivationCIVIL-226: Introduction to machine learning for engineers

Explores neural networks, activation functions, backpropagation, and PyTorch implementation.

Convolutional Neural NetworksME-390: Foundations of artificial intelligence

Covers convolutional neural networks, filter operations, and their applications in signal processing and image analysis.

Document Analysis: Topic ModelingDH-406: Machine learning for DH

Explores document analysis, topic modeling, and generative models for data generation in machine learning.