Skip to content

Data Science

  • BERT
    • ML Framework for NLP. It helps computer to understand a language and context of text by using surrounding text.
    • Bidirectional Encoder Representations from Transformers
  • Generative design is a technology in which 3D models are created and optimized by cloud computing and AI. A user sets up requirements for the model, such as manufacturing processes, loads, and constraints, and then the software offers designs that meet those requirements.

Table of Index Kaggle

Maths:

Data ETL EDA Wrangling

Regression - supervised:

  • Linear Regression
  • Simple Linear Regression
  • Multiple Linear Regression
  • Polynomial Linear Regression
  • Support Vector Regression
  • Decision Tree Regression
  • Random Forest Regression

Classification - supervised:

  • Logistic Regression - confusion matrix, accuracy, sigmoid, CAP Curve
  • KNN - K Nearest Neighbour Classifier - Euclidean distance
  • SVM - Support Vector Machines - Maximum Margin Hyperplane
  • Kernel SVM - Map to Higher Dimention, Kernel Trick, Gaussian RBF Kernel
  • Naive Bayes - Bayes Theorem
  • Decision Tree Classifier
  • Random Forest Classifier - entropy, ensemble learning

Clustering - unsupervised:

  • K-Means Clustering - wcss, choosing k-value, k-means++
  • Hierarchical Clustering - Agglomerative, Dendrogram

Association Rule - Unsupervised:

  • Apriori Algorithm - market basket analysis, lift, support, confidence, todo
  • Eclat (Part 5 todo)
  • FP Growth

Other Bonus Extra:

  • Model Select
  • Reinforcement Learning (Part 6 todo)
    • Upper Confidence Bound
    • Thompson Sampling

NLP:

  • Cleaning, stemming, nltk, bag of words, tokenization

Deep Learning: (Part 8 todo)

Dimentionality Reduction:

  • PCA - max variance, unsupervised
  • LDA - max class separation, supervised
  • Kernel PCA - kernel trick on non-linearly separable dataset
  • SVD
  • GDA - Generalized Discriminant Analysis
2025-01-12 May 2021