Course Outline
Machine Learning Algorithms in Julia for Government
Introductory Concepts
- Supervised and unsupervised learning
- Cross-validation and model selection
- Bias/variance tradeoff
Linear and Logistic Regression
(NaiveBayes & GLM)
- Introductory concepts
- Fitting linear regression models
- Model diagnostics
- Naive Bayes
- Fitting a logistic regression model
- Model diagnostics
- Model selection methods
Distances
- What is a distance?
- Euclidean
- Cityblock
- Cosine
- Correlation
- Mahalanobis
- Hamming
- Median Absolute Deviation (MAD)
- Root Mean Square (RMS)
- Mean Squared Deviation
Dimensionality Reduction
-
Principal Component Analysis (PCA)
- Linear PCA
- Kernel PCA
- Probabilistic PCA
- Independent Component Analysis (ICA)
- Multidimensional Scaling
Altered Regression Methods
- Basic concepts of regularization
- Ridge regression
- Lasso regression
- Principal Component Regression (PCR)
Clustering
- K-means
- K-medoids
- DBSCAN
- Hierarchical clustering
- Markov Cluster Algorithm
- Fuzzy C-means clustering
Standard Machine Learning Models
(NearestNeighbors, DecisionTree, LightGBM, XGBoost, EvoTrees, LIBSVM packages)
- Gradient boosting concepts
- K nearest neighbors (KNN)
- Decision tree models
- Random forest models
- XGBoost
- EvoTrees
- Support vector machines (SVM)
Artificial Neural Networks
(Flux package)
- Stochastic gradient descent and strategies
- Multilayer perceptrons: forward feed and back propagation
- Regularization
- Recurrent neural networks (RNN)
- Convolutional neural networks (Convnets)
- Autoencoders
- Hyperparameters
Requirements
This course is designed for individuals who already possess a background in data science and statistics, aligning with the advanced training needs for government professionals.
Testimonials (2)
the ML ecosystem not only MLFlow but Optuna, hyperops, docker , docker-compose
Guillaume GAUTIER - OLEA MEDICAL
Course - MLflow
I enjoyed participating in the Kubeflow training, which was held remotely. This training allowed me to consolidate my knowledge for AWS services, K8s, all the devOps tools around Kubeflow which are the necessary bases to properly tackle the subject. I wanted to thank Malawski Marcin for his patience and professionalism for training and advice on best practices. Malawski approaches the subject from different angles, different deployment tools Ansible, EKS kubectl, Terraform. Now I am definitely convinced that I am going into the right field of application.
