MLintro-HW-Notes-Projects

Special Notes

• This doc contains the illustrations for the notebooks (notes of mine), but it's still RECOMMENDED to read teacher's OFFICIAL NOTES as it contains more detailed and explained information
• Most of the materials are come from Udemy python for DS and ML bootcamp. Please refer to its term of service and course behavior rule when you want to use anything from this repo.

---

s15

s16-Bias Variance Trade-off-Notes

• bias variance trade-off definition & significance
• varaince & bias definition, and their relationships
• overfit & underfit
• causes of high bias / variance
• how to reduce variance in a model

s17-Logistic-Regression-theory-Notes

• logistic regression
• pandas.DataFrame.isnull and seaborn.heatmap to clean the data with missing values
• use get_dummies to clean the data
• use classification_report/confusion matrix to evaluate the model (first)
• how to increase precision / recall / accuracy?

s18-K-Nearest-Neighbors-Notes

• KNN Classification definition, pros/cons, how it works
• use of StandardScaler to standarlize the data
• use of elbow method to determine the optimal k value for KNN classification
• train, evaluate the model

s19-Decision-Trees-and-Random-Forests-Notes

• Decision Tree / Random Forest Classifier definition, when to use, how they work
• Decision Tree concepts.
• Random Forest: Pros/Cons, when to use...
• specify parameter n_estimator to specify number of trees
• POST: random forest / decision tree explained by Jose

s20-Support-Vector-Machines-Notes

• Support Vector Machine definition, when to use, how they work, concepts(regression analysis, support vectors, hyperplane)
• Kernal Trick
• Breast_cacer dataset from sklearn
• Grid Search method: takes a dictionary object and searches the best combination of parameters values by inspecting the cross-validation.
• PROJECT: using Iris dataset from sklearn to classify flowers.

s21-K-Means-Clustering-Notes

• KMeans definition (kw. Unsupervised, cluster, centroid, elbow method) when to use, how works, how to use
• make_blobs method from sklearn.datasets (generate data in clusters with 'labels')
• plt.subplot to compare results from dataset and kmeans results
• PROJECT: using Real college dataset (generate clusters and) to predict whether is private. (kw. sns.FacetGrid)

s22-Principal-Component-Analysis-Notes

• PCA's definition, (kw. Unsupervised, data analysis technique, interrelationships of vairables, factor analysis, dimension reduction) how to use, when to use
• Get PCA's components
• REAPPEAR: StandardScaler method and breast_cancer dataset
• NO PROJECT, reading teacher's notes is recommended

s23-Recommender-Systems-Notes

• Recommender System's catogories, definitions (kw. Content-Based, Collaboritive-Filtering, complex), how to use, when to use
• pandas.DataFrame.pivot_table: to organize the data from DataFrame
• pandas.DataFrame.corrwith: get the correlation(0-1) between two DataFrames
• REAPPEAR: pandas merge, groupby, sort_values, jointplot...
• NO PROJECT, also reading teacher's notes is recommended (advanced RS)