Content of course: "Elements of Artificial Intelligence - Lectures"
1. Introduction. Fundamentals of Artificial Intelligence and Deep Learning. Trainable machines, types of training
2. Statistical models. Regression types: simple, multiple, polynomial. Logistic regression. Cost function. The gradient descent (GD) method for trainable machines. Derivatives. GD for logistic regression. GD for multiple training examples
3. Aspects of logistic regression implementation from the perspective of neural networks, using Python. Case study.
4. Artificial Neural Networks (ANNs) – fundamental concepts. Artificial neuron. Artificial Neural Networks. with Forward Propagation. Output determination. Activation functions. Descent gradient for ANN. Backpropagation.
5. Aspects of ANN implementation, using Python. Forward and backward propagation. Case study.
6. L-Layer Deep Neural Networks. Structure, building blocks. Forward and backward propagation. Parameters and hyperparameters.
7. Aspects of deep neural network implementations, using Python. Network structure; parameter initialization; forward propagation; the loss function; backward propagation; parameter optimization; usin the trained network
8. Improving Deep Neural Networks: Hyperparameter optimization; regularization; normalization; optimization algorithms
9. Convolutional neural networks. Structure. Edge detection; 2D convolution; convolution layers; aggregation layers; fully connected layers
10. Aspects regarding the implementation of convolutional networks in Python using TensorFlow
11. Classic convolutional neural networks: LeNet ‐ 5; AlexNet; VGG‐16; ResNets; Inception network. Data augmentation.
12. Detection algorithms. Object location, object detection, prediction of bounding rectangles, intersection/union (IoU), suppression of non‐maxima, anchor rectangles.
13. The YOLO algorithm. Vehicle detection using YOLO
14. Face recognition. Siamese convolutional networks. Checking and classification of faces. Style transfer for image generation, using neural networks
Content of lab: "Elements of Artificial Intelligence - Lab"
1. Google Collaboratory. Notebook. Using of Python in Colab
2. Introduction to Python. Sigmoidal function, ReLU function, massive data size changing, normalization, broadcasting, softmax function, vectorization, loss functions
3. Planar data classification with ANN
4. Implementation of the deep neural network for binary classification applications
5. Convolutional neural network for object recognition - CIFAR10
6. Vehicle detection using YOLO. Counting vehicles using YOLO
7. Synthesis laboratory
