Each class will consist of about 60 minutes of lecture followed by demos inspired by real-world applications
for the next 20 minutes. These demos will be shown by the instructor using MATLAB (and may also include
some embedded systems for tangible “real-world” applications) so that the students can visualize the practical
applications of the concepts taught in the class.

Week | Topics |
---|---|

Week 1 |
What is Machine Learning (ML)? How is it different from traditional programming? Why use ML? Types of ML systems Feature Selection and Extraction on multimodal data |

Week 2 |
Feature Preprocessing (cleaning, handling missing values, one-hot encoding) Dividing dataset into train/ validation/ test sets Basic Pattern Recognition Tools on Extracted Features (Template-based Matching, Correlation, Covariance, HOG features) |

Week 3 |
Features Clustering k-Means, Finding optimal k (number of classes) for clustering GMM Clustering and DBScan Clustering methods |

Week 4 |
Features Dimensionality Reduction Curse of Dimensionality, Principal Component Analysis (PCA), Independent Component Analysis (ICA), Linear Discriminant Analysis (LDA) [Quiz 1] |

Week 5 |
Features Classification and Regression I Distance-based Classification, k-Nearest Neighbors Classification Performance Metrics (Confusion Matrix, ROC Curve, F1-score vs. Accuracy, Precision, Sensitivity, etc.) and Data Sampling |

Week 6 |
Features Classification and Regression II SVMs, Kernels, Support Vector Regression, Voting Classifier Semi-supervised Learning |

Week 7 |
Main challenges of ML Systems (insufficient data, overfitting, poor quality data,
irrelevant features, noise in the dataset, etc.) Responsible AI: Fairness, Inclusion, and Ethics in ML (Biases and Examples) |

Week 8 | [Midterm Project Evaluation] |

Week 9 |
Performing Computer Vision, Time-series, and NLP Analyses with above techniques on real-world data (applications in Computer Science, Finance, Biology, and Robotics) [Quiz 2] |

Week 10 |
Introduction to Neural Networks (NNs) Biological NNs and parallels with Artificial NNs, Mathematical formalization of neurons (Perceptron), Perceptron Training algorithm, Learning and linear separability in a Perceptron, XOR Problem Multilayer Neural Networks, Backpropagation and its training algorithm, solving the XOR problem, decision boundaries by hidden layers |

Week 11 |
Hyperparameters Tuning in Neural Networks Learning rate, Momentum, Adaptive Learning Rate, Accelerated Learning, Weights Initialization (vanishing and exploding gradients, and dead neurons) Regularization (L1, L2, and Dropout), Batch Normalization, Mini-batch |

Week 12 |
Deep Learning (Visual Data) Convolutional Neural Networks (CNN), Common CNN Architectures Overview, Main Concepts: Weight-sharing, Convolution, Padding, Pooling CNN Spatial Dropout, BatchNorm, deeper dive into the common CNN Architectures (VGG-16, ResNets, NiN, GoogLeNet, Inception v3, etc.), Transfer Learning |

Week 13 |
Recurrent Neural Networks Backpropagation through time, Long-short term memory (LSTM), Many-to-one word RNNs, Generating text with character RNNs Attention Mechanisms and Transformers |

Week 14 |
Generative Adversarial Networks (GANs) Autoencoders, Fully-connected Autoencoders, Convolutional Autoencoders GANs loss function, applications and different types of GANs |

Week 15 |
[Quiz 3] Genetic Algorithms, Reinforcement Learning |

Week 16 | [Final Project Evaluation] |