Full Stack Machine Learning and AI Program
This program offers comprehensive learning in applying Machine Learning, Deep Learning and Artificial Intelligence techniques to both structured and unstructured data including text, images, audio and video data. After this course, you will be able to select, apply and deploy appropriate Machine Learning and Deep Learning algorithms depending on the type of problems and data involved. The course provides a substantial intuition of the mathematics behind each of these algorithms, as well as business applications in AI using real life case studies.
24 weeks
10 hours / week
12 months
Modules
Python for Data Analysis, Machine Learning Algorithms in Python and Deep Learning Algorithms
Tools
Caret, Dplyr, Ggplot2, Keras, Numpy, OpenCV, Pandas, Python, R, Scipy, Sklearn, TensorFlow
MODULES INCLUDED
- Introduction to Machine Learning
- Types of ML Algorithms: Supervised and Unsupervised learning
- Common ML Tasks: Regression, Classification
- Introduction to Python: Set up using Anaconda
- Basic Python Data Structures - Integer, Floats,Strings, Booleans,Lists, Tuples, Dictionary, Files
- Functions and Methods, Iterator and Itearables, List Comprehension, Map and Lambda.
- Introduction to Pandas: Pandas Series, Data Frames, Importing Data
- Data Manipulation with Pandas: Filtering, Selecting, Sorting, Adding new columns, Group by operations
- Handling Strings, Dates
- Combining Dataframes using pandas
- Visualization in Python: Creating Univariate and Bivariate Plots using MatplotLib and pandas
- Subplots, Figure vs. Axes in MatplotLib
- Univariate and Faceted plots using seaborn.
- Reading Images in Python using MatpotLib
- Case Study: Manipulate Retail Store Orders Data
- Case Study: Visualize on HR Data
- Case Study: Analyse World Terror incidents data
- An Introduction to Feature Engineering and its scope
- An Overview of Machine Learning Pipeline using CRISP-DM Methodology
- Feature Understanding : Summary, missing value and outlier treatment
- Feature Engineering One Hot Encoding/ Interaction Features/Normalization/Discretisation: Create indicator/ dummy(One Hot Encoding) variables
- Feature Selection: attributes vs. features
- Principal Component Analysis (PCA) on structured data
- Feature Engineering on Text and Image Data: Import flat files, images,tf-idf, grayscale images, Need of resizing images, import images from multiple folders
- Count vector representation,tf-idf, cosine similarity (unigram, bigram, trigrams)
- Managing tfidf using high dimensionality reduction techniques like SVD and NMF
- Feature Engineering on Image Data
- Import images, Converting images to grayscale, resizing images
- Case Study: Prepare Credit Default (structured) data for ML Algorithms
- Case Study: Engineer Amazon Product Reviews - Text Data
- Case Study : Engineer Handwritten Digit Images
- Introduction to Linear Regression and OLS
- Finding Model Parameters using Gradient Descent and Stochastic Gradient Descent
- Concept of Overfitting, Bias-Variance Tradeoff and Regularisation
- Parameter Tuning- Grid Search using K-Fold CV
- Logistic Regression - Binary and Multiclass Classification
- Understanding Logistic Cost Function
- Measuring Classification Performance
- Case Study: Predict Online News Popularity
- Case Study: Sentiment Analysis of Movie Reviews
- Case Study : Detect handwritten digits
- Introduction to Classification Trees and Regression Trees
- Purity Metrics - Gini, Entropy, MSE
- Parameter Tuning- Grid Search using K-Fold CV
- Tree Based Ensembles - Bagged Trees and Random Forests Algorithms
- Hyperparamater Tuning using OOB score
- Feature Importance using Tree Based Ensemble Models
- Adaboost and Gradient Boosting
- Partial Dependence Plots
- Case Study: HR Churn/ Employee Attrition prediction
- Case Study: predict Airbnb property Prices
- Case Study: Analysing Customer Behavior in Direct Marketing
- Memory Based Engines
- Item Based Collaborative Filtering: Similarity measures
- User Based Collaborative Filtering
- Similarity and Accuracy Measures
- Model Based Engines: SVD & NMF(Matrix Factorization)
- Hyperparamater Tuning using GridSearch CV
- Case Study: Build a song recommendation engine
- Introduction to Clustering - Kmeans and Hierarchical
- Data Preparation for Clustering - Scaling and Weighting
- Deciding optimum number of clusters: Elbow Method/Scree Plot, Silhouette Measure
- Creating cluster profiles
- Agglomerative Clustering: Linkages/ Dendogram
- Case Study: Segment San Francisco Housing Data
- Case Study: Segment Football Players
- Case Study: Image Segmentation based on Cigarette Brands
- Case Study: Segment various States in India and US based on Pollution Levels
- Introduction to Neural Networks: Scope of Neural Networks
- Multi Layered Perceptrons (MLP)
- Activation Functions : ReLU, Sigmoid, tanh, Softmax
- Model Training: Forward Pass, Backward Pass
- Dropout Layers: Avoid Overfitting
- Convolutional Neural Network(CNN)
- Convolutional Layers: kernel , padding, input shape
- Pooling Layers: pooling size
- Fully connected Layers: dense
- Implementing MLP & CNN using Keras with TensorFlow as backend
- Case Study: Image Recognition on Digits data
- Case Study: Image Recognition on Fashion data
- What is Blockchain?, What is the mechanism involved?
- Transforming the Trade Finance Process
- Features and Vocabulary
- Vocabulary, Benefits and Challenges
- Demystifying Cryptocurrency and Smart Contracts
- Blockchain Then and Now
- Blockchain 1.0,2.0,3.0
- Use Cases Driven by Blockchain
- Introduction to R: Import Flat files
- Manipulating Data with R: summaries, aggregation, missing and outlier treatment
- TM Package in R: Corpus, clean text, Term document matrix
- Machine Learning in R: Implementing Supervised and Unsupervised Machine Learning, Tree Based models, Ensembles, Neural Networks and CNN in R
- Case study codes for R implementations
- Project: Build a CNN to detect regional languages using Devanagari digits.
- libraries to be used: glob,cv2, Keras with TensorFlow in the backend.
- Comments: Need to import images from multiple sub folder from a given directory and should be tagged properly after extracting pixel intensities and framing them in a format which any machine learning algorithm can understand
- Build Your Own Chatbot
- Deep Learning with TensorFlow
LEARN FROM THE EXPERTS
Gunnvant Singh
Gunnvant has 3+ years of teaching experience and is an expert in business statistics, finance and data science with R.
Neha Shitut
Neha is an advanced Analytics ML expert with diverse professional analytics experiences. She has spent 6 years working with Genpact in Retail Analytics primarily and last 5 years she has been teaching Data Science in R, Python and SAS. A
career facts
India – INR 6 lakhs to 12 lakhs (entry level)
Global – $65,000 to $130,000
Works closely with product teams to design and implement data based products; designs solutions for business problems leveraging data
Lead Data Scientist or Chief Data Scientist
Machine Leaning Specialist or Lead ML Specialist
Retail, eCommerce, Banks, Telecom, FMCG, Manufacturing and other domains where data is generated.
WHAT YOU GET
The Catalyst Approach
Learn through video lectures, live Q&A sessions conducted by the faculty, assignments, interactive case study workshops with senior faculty and industry mentors. Know more.
Jigsaw Learning Center
Gain free access to a variety of resources like handouts, reference material, guides, lecture transcripts and student forums for your course duration.
In-person Faculty Support
Get all your course-related queries solved by the Jigsaw Faculty in-person for the program duration.
General Support
We have your back! Our student support teams are available on the weekends for call and on email for your course duration.
Q&A Online Sessions
Have questions? Ask our analytics mentors during the 2 hour Q&A session every Saturday for your program duration.
Career Counseling
Avail professional guidance on resume building, interview preparation and identification of relevant opportunities, for the analytics field.
Mobile App
Learn on the go, wherever you are with mobile access to Jigsaw's course videos.
Access to Jigsaw Lab
Get access to a cloud-based solution, for hands-on experience with real-life business data using the latest analytical tools. The access tools include SAS, R, Hadoop, MapReduce, Hive, Pig, Sqoop etc.
Capstone Project
Hands-on industry project experience
FINAL OUTCOME
After completing this Machine Learning and AI certification, you would have mastered the most popular and powerful Machine Learning and Deep Learning algorithms that are commonly used across multiple industries to solve large scale problems with data, and are also used in building AI systems.
If you need clarity on how online training works, please visit
Or if you have any other questions, please visit the
In case you still have any unanswered questions, we encourage you to register for an upcoming webinar.
