This is a summary of the the Data Science Diploma from Concordia.
https://concordiabootcamps.ca/courses/data-science-remote/
The jupyter notebooks are a mix of theory and useful python code intented to be used as a quick reference.
-
Python
- Data structures
- Functions
- Strings
- Regex
- Input / Output
- Classes
- Numpy
- Magic commands
-
Algorithms
- Sorting
- Recursion
- Dynamic Programming
- Graphs
- Fast code
-
Pandas
- Importing
- Explore
- Slicing
- Filtering
- Ploting
- Group by
- Merging
- Strings
- JSON
- Time series
-
Visualization
- Matplotlib
- Seaborn
- Examples
-
Regression
- Simple regression
- Multiple regressions
- Matrix form regression
- Interpretation
- Feature and target engineering
- Regularisation
-
Classification
- Root finding
- Optimization
- Logistic regression
- Interpretation
- ROC curve
- Exotic distributions
- Two stage modeling
- Survival model
-
SQL
- Connect
- Query
- Group by
- Join
- Nested queries
- Create table
- Create DB
-
Clustering
- K-Means
- Spectral Clustering
- Image compression
- Metrics
- Model evaluation
- Selecting the number of clusters
-
Scraping
- API requests
- BeautifulSoup
- Selenium
-
Dimentionality Reduction (Embeddings)
- PCA
- UMAP
- T-SNE, MDA, etc.
-
NLP
- Tokenisation
- Stemming
- Lemmatization
- Stop Words
- Matching
- Name Entity Regognition
- Features extraction
- Word Vectors (embeddings)
- Sentiment Analysis
- Topic Modeling (LDA, NMF)
- Summarization
-
ML Models
- SVM
- Decision Tree
- Gradient Boosting
- Shapley
-
Deep Learning
- FFNN
- CNN
- RNN: GRU, LSTM
- RL: GAN
-
Time Series
- ARIMA models
- VAR models
- Panels
-
Model Deployment
- AWS
- GCP
-
ML Tools
- Scaling, Normalizing
- Polynomial features
- Test split
- Cross Validation
- Grid Search
- Pipeline
-
Boilerplate (to come)
- SKLearn
- TensorFlow (Keras)
- PyTorch
-
Recommender System (to come)
- Collaborative Filtering
- Content-based Filtering
- Hybrid
-
Other (to come)
- Computer Vision
- Naive Baye? (Classifier)
- Markov's chain? (NN)
Useful?
