Data Science (Machine Learning)

Detailed Syllabus

Introduction to Data Science

• What is Data Science
• What is Machine Learning
• What is Deep Learning
• Role of Data Scientist
• Applications of Data Science
• Data and its sources
• Overview of Data Science Life Cycle

Working with Anaconda and Jupyter Notebook

• Downloading and installing Anaconda
• Starting Jupyter Notebook
• UI elements of Notebook
• Kernel and types of cells - Code and Markdown
• Modes - Edit and Command
• Magic functions - Line and Cell functions
• Keyboard shortcuts - Command mode and Edit mode shortcuts
• Saving and loading of notebook
• Using Jupyter Lab

Basic Statistics

• Mean, Median, Mode and Range
• Using statistics module
• Variance and Standard Deviation
• Quartiles and IQR
• Understanding distribution of data using Histogram and Box plot
• Measuring Skewness and Kurtosis
• Probability
• Correlation between variables
• Using scipy.stats module
• Scatter plot to understand correlation
• Regression Analysis
• Understanding intercept and slope - predict y given X

Numpy

• Creating single and multi-dimensional arrays
• Using indexing and slicing
• Using fancy indexing (boolean indexing and array of indices)
• Array operations, methods of ndarray and universal functions
• View vs. Copy of array
• Reshaping arrays
• Stacking and splitting arrays
• Broadcasting
• Applying Linear Algebra
• Image processing with Arrays

Pandas

• Working with Series
• Applying methods on Series
• Working with DataFrame
• Reading data into DataFrame and writing DataFrame to other formats
• Selecting rows and columns in DataFrame
• Adding and deleting rows and columns in DataFrame
• Working with apply() and applymap() functions
• Working with str attribute for string manipulations
• Joining, Merging and Concatenating DataFrames
• Grouping data on one or more columns
• Using pivot_table()
• Data Wrangling - Binning, Encoding etc.
• Handling null values
• Drawing plots using Pandas

Matplotlib

• Anatomy of a figure
• Working with Module API and Object API
• Working with different plots - Histogram, Bar, Stacked Bar, Pie, Scatter, Line
• Creating multiple axes in single figure
• Customizing plots - labels, legends, scales, titles, text etc.

Seaborn

• Figure-level vs. Axes level plots
• Categorical, Relational, Distribution, Regression and Matrix Plots
• Using parameters like hue, row and col

Data Science Workflow (Life Cycle) using Classification Case Study

• What is the question
• Data Acquisition
• Preparing data - cleaning and organizing data
• Exploratory Data Analysis (EDA)
• Data Munging/Data Wrangling
• Feature Engineering
• Data Visualization
• Model Building
• Model Evaluation
• Model Deployment

Machine Learning Workflow with Classification Case Study

• Understanding pre-processing concepts like Standardization, Encoding etc.
• Training Model using train and test split
• Using different algorithms like Logistic Regression, Support Vector Machines, k-Nearest Neighbors, Naive Bayes, Decision Tree, Random Forest using Scikit-learn
• Evaluating result of the model using metrics - classification report, confusion matrix
• Understand Precision, Recall, F1 Score, Specificity and Sensitivity
• Understanding cross validation and how to use it to train and test model
• Presenting the model - Deployment of the model

Working with Regression case study

• How to use metrics - MSE, RMSE, R2 Score etc. to evaluate model
• Understanding Regularization - Lasso and Ridge
• Understanding ensemble algorithms - Bagging and Boosting
• Stochastic Gradient Descent
• Using Grid Search to select right hyper parameters
• How to use Pipeline
• Understanding non-linear data - polynomial features

Unsupervised Machine Learning - Clustering and Association

• What is clustering
• How k-Means clustering works
• How DBSCAN works to create clusters
• How hierarchical clustering works - Agglomerative clustering and Dendrogram
• Recommender systems - Collaborative filtering and Content-based filtering