Title: Implementing AI Algorithms from Scratch

Location: US

Company: CodeSignal

About this course:
This advanced course path is for learners who want to move beyond using machine learning libraries and develop a deeper understanding of how foundational AI algorithms are implemented. You will build key models and methods directly, focusing on the mechanics behind training, prediction, optimization, and evaluation. The path begins with neural network fundamentals, including perceptrons, activation functions, and multi-layer network components. It then expands into unsupervised learning, ensemble methods, classification, regression, and gradient-based optimization techniques. By implementing algorithms such as k-means, PCA, random forests, boosting methods, logistic regression, decision trees, and Adam optimization from scratch, you will strengthen both your mathematical intuition and your practical programming skills for AI development.

Implement neural network components such as perceptrons, activation functions, and multi-layer structures

Build unsupervised learning algorithms including k-means, mini-batch k-means, PCA, and DBSCAN

Construct ensemble methods such as bagging, random forests, AdaBoost, and gradient boosting

Develop optimization algorithms including stochastic gradient descent, momentum, RMSProp, and Adam

Create classification models such as logistic regression, k-nearest neighbors, Naive Bayes, and decision trees

Evaluate models and clusters using metrics such as AUC-ROC, homogeneity, completeness, and v-measure

Implement simple and multiple linear regression with gradient descent

Build core artificial intelligence and machine learning algorithms from scratch to understand how they work beneath high-level libraries. This advanced path covers neural networks, clustering, ensembles, optimization, classification, and regression.

Duration: 7 weeks

Curriculum:
- Implement neural network components such as perceptrons, activation functions, and multi-layer structures
- Build unsupervised learning algorithms including k-means, mini-batch k-means, PCA, and DBSCAN
- Construct ensemble methods such as bagging, random forests, AdaBoost, and gradient boosting
- Develop optimization algorithms including stochastic gradient descent, momentum, RMSProp, and Adam
- Create classification models such as logistic regression, k-nearest neighbors, Naive Bayes, and decision trees
- Evaluate models and clusters using metrics such as AUC-ROC, homogeneity, completeness, and v-measure

- Implement simple and multiple linear regression with gradient descent