Machine Learning Certification Training using Python

Goal: GDiscuss the different sources available to extract data, arrange the data in structured form, analyze the data, and represent the data in a graphical format.

Objectives: At the end of this Module, you should be able to:

• Discuss Data Acquisition techniques
• List the different types of Data
• Evaluate Input Data
• Explain the Data Wrangling techniques
• Discuss Data Exploration

Topics:

• Data Analysis Pipeline
• What is Data Extraction
• Types of Data
• Raw and Processed Data
• Data Wrangling
• Exploratory Data Analysis
• Visualization of Data

Hands-On/Demo:

• Loading different types of dataset in Python
• Arranging the data
• Plotting the graphs

Goal: In this module, you will learn the concept of Machine Learning and it’s types.

Objectives: At the end of this Module, you should be able to:

• Essential Python Revision
• Necessary Machine Learning Python libraries
• Define Machine Learning
• Discuss Machine Learning Use cases
• List the categories of Machine Learning
• Illustrate Supervised Learning Algorithms
• Identify and recognize machine learning algorithms around us
• Understand the various elements of machine learning algorithm like parameters, hyper parameters, loss function and optimization.

Topics:

• Python Revision (numpy, Pandas, scikit learn, matplotlib)
• What is Machine Learning?
• Machine Learning Use-Cases
• Machine Learning Process Flow
• Machine Learning Categories
• Linear regression
• Gradient descent

Hands-On/Demo:

• Linear Regression – Using Boston Dataset

Goal: In this module, you will learn Supervised Learning Techniques and their implementation, for example, Decision Trees, Random Forest Classifier etc.

Objectives: At the end of this Module, you should be able to:

• Understand What is Supervised Learning?
• Illustrate Logistic Regression
• Define Classification
• Explain different Types of Classifiers such as Decision Tree and Random Forest

Topics:

• What is Classification and its use cases?
• What is Decision Tree?
• Algorithm for Decision Tree Induction
• Creating a Perfect Decision Tree
• Confusion Matrix
• What is Random Forest?

Hands-On/Demo:

• Implementation of Logistic regression, Decision tree, Random forest

Goal: In this module you will learn about impact of dimensions within data. You will be taught to perform factor analysis using PCA and compress dimensions. Also, you will be developing LDA model.

Objectives: At the end of this Module, you should be able to:

• Define the importance of Dimensions
• Explore PCA and its implementation
• Discuss LDA and its implementation

Topics:

• Introduction to Dimensionality
• Why Dimensionality Reduction
• PCA
• Factor Analysis
• Scaling dimensional model
• LDA

Hands-On/Demo:

• PCA
• Scaling

Goal: In this module, you will learn Supervised Learning Techniques and their implementation, for example, Decision Trees, Random Forest Classifier etc.

Objectives: At the end of this Module, you should be able to:

• Understand What is Naïve Bayes Classifier
• How Naïve Bayes Classifier works?
• Understand Support Vector Machine
• Illustrate How Support Vector Machine works?
• Hyperparameter optimization

Topics:

• What is Naïve Bayes?
• How Naïve Bayes works?
• Implementing Naïve Bayes Classifier
• What is Support Vector Machine?
• Illustrate how Support Vector Machine works?
• Hyperparameter optimization
• Grid Search vs Random Search
• Implementation of Support Vector Machine for Classification

Hands-On/Demo:

• Implementation of Naïve Bayes, SVM

Goal: In this module, you will learn about Unsupervised Learning and the various types of clustering that can be used to analyze the data.

Objectives: At the end of this Module, you should be able to:

• Define Unsupervised Learning
• Discuss the following Cluster Analysis

Topics:

• What is Clustering & its Use Cases?
• What is K-means Clustering?
• How K-means algorithm works?
• How to do optimal clustering
• What is C-means Clustering?
• What is Hierarchical Clustering?
• How Hierarchical Clustering works?

Hands-On/Demo:

• Implementing K-means Clustering
• Implementing Hierarchical Clustering

Goal: In this module, you will learn about Unsupervised Learning and the various types of clustering that can be used to analyze the data.

Objectives: At the end of this Module, you should be able to:

• Define Association Rules
• Learn the backend of recommendation engines and develop your own using python

Topics:

• What are Association Rules?
• Association Rule Parameters
• Calculating Association Rule Parameters
• Recommendation Engines
• How Recommendation Engines work?
• Collaborative Filtering
• Content Based Filtering

Hands-On/Demo:

• Apriori Algorithm
• Market Basket Analysis

Goal: In this module, you will learn about developing a smart learning algorithm such that the learning becomes more and more accurate as time passes by. You will be able to define an optimal solution for an agent based on agent environment interaction.

Objectives: At the end of this Module, you should be able to:

• Explain the concept of Reinforcement Learning
• Generalize a problem using Reinforcement Learning
• Explain Markov’s Decision Process
• Demonstrate Q Learning

Topics:

• What is Reinforcement Learning
• Why Reinforcement Learning
• Elements of Reinforcement Learning
• Exploration vs Exploitation dilemma
• Epsilon Greedy Algorithm
• Markov Decision Process (MDP)
• Q values and V values
• Q – Learning
• α values

Hands-On/Demo:

• Calculating Reward
• Discounted Reward
• Calculating Optimal quantities
• Implementing Q Learning
• Setting up an Optimal Action

Goal: In this module, you will learn about Time Series Analysis to forecast dependent variables based on time. You will be taught different models for time series modelling such that you analyse a real time dependent data for forecasting.

Objectives: At the end of this Module, you should be able to:

• Explain Time Series Analysis (TSA)
• Discuss the need of TSA
• Describe ARIMA modelling
• Forecast the time series model

Topics:

• What is Time Series Analysis?
• Importance of TSA
• Components of TSA
• White Noise
• AR model
• MA model
• ARMA model
• ARIMA model
• Stationarity
• ACF & PACF

Hands-On/Demo:

• Checking Stationarity
• Converting a non-stationary data to stationary
• Implementing Dickey Fuller Test
• Plot ACF and PACF
• Generating the ARIMA plot
• TSA Forecasting

Goal: In this module, you will learn about selecting one model over another. Also, you will learn about Boosting and its importance in Machine Learning. You will learn on how to convert weaker algorithms to stronger ones.

Objectives: At the end of this Module, you should be able to:

• Discuss Model Selection
• Define Boosting
• Express the need of Boosting
• Explain the working of Boosting algorithm

Topics:

• What is Model Selection?
• Need of Model Selection
• Cross – Validation
• What is Boosting?
• How Boosting Algorithms work?
• Types of Boosting Algorithms
• Adaptive Boosting

Hands-On/Demo:

• Cross Validation
• AdaBoost

Goal: In this module, you will learn how to approach and implement a Project end to end, and a Subject Matter Expert will share his experience and insights from the industry to help you kickstart your career in this domain. Finally, we will be having a Q&A and doubt clearing session.

Objectives: At the end of this Module, you should be able to:

• How to approach a project
• Hands-On project implementation
• What Industry expects
• Industry insights for the Machine Learning domain
• QA and Doubt Clearing Session

Data Science is a set of techniques that enables the computers to learn the desired behavior from data without explicitly being programmed. It employs techniques and theories drawn from many fields within the broad areas of mathematics, statistics, information science, and computer science. This course exposes you to different classes of machine learning algorithms like supervised, unsupervised and reinforcement algorithms. This course imparts you the necessary skills like data pre-processing, dimensional reduction, model evaluation and also exposes you to different machine learning algorithms like regression, clustering, decision trees, random forest, Naive Bayes and Q-Learning.

After completing this Machine Learning Certification Training using Python, you should be able to:

• Gain insight into the 'Roles' played by a Machine Learning Engineer
• Automate data analysis using python
• Describe Machine Learning
• Work with real-time data
• Learn tools and techniques for predictive modeling
• Discuss Machine Learning algorithms and their implementation
• Validate Machine Learning algorithms
• Explain Time Series and it’s related concepts
• Gain expertise to handle business in future, living the present

Brighter Connect’s Python Machine Learning Certification Course is a good fit for the below professionals:

• Developers aspiring to be a ‘Machine Learning Engineer'
• Analytics Managers who are leading a team of analysts
• Business Analysts who want to understand Machine Learning (ML) Techniques
• Information Architects who want to gain expertise in Predictive Analytics
• 'Python' professionals who want to design automatic predictive models

The pre-requisites for the Machine Learning Certification Training using Python includes development experience with Python. Fundamentals of Data Analysis practised over any of the data analysis tools like SAS/R will be a plus. However, Python would be more advantageous. You will be provided with complimentary “Python Statistics for Data Science Course” as a self-paced course once you enrol for the course.

You will do your Assignments/Case Studies using Jupyter Notebook that is already installed on your Cloud LAB environment whose access details will be available on your LMS. You will be accessing your Cloud LAB environment from a browser. For any doubt, the 24*7 support team will promptly assist you.

This course comprises of 34 case studies that will enrich your learning experience. In addition, we also have 3 Projects that will enhance your implementation skills. Below are few case studies which are part of this course:

Case Study 1: Maple Leaves Ltd is a start-up company which makes herbs from different types of plants and its leaves. Currently the system they use to classify the trees which they import in a batch is quite manual. A laborer from his experience decides the leaf type and subtype of plant family. They have asked us to automate this process and remove any manual intervention from this process.

You have to classify the plant leaves by various classifiers from different metrics of the leaves and to choose the best classifier for future reference.

Case Study 2: BookRent is the largest online and offline book rental chain in India. Company charges a fixed fee per month plus rental per book. So, company makes more money when user rent more books.

You as an ML expert and must model recommendation engine so that user gets recommendation of books based on behavior of similar users. This will ensure that users are renting books based on their individual taste. .

Company is still unprofitable and is looking to improve both revenue and profit. Compare the Error using two approaches – User Based Vs Item Based .

Case Study 3: Handle missing values and fit a decision tree and compare its accuracy with random forest classifier.

Predict the survival of a horse based on various observed medical conditions. Load the data from ‘horses.csv’ and observe whether it contains missing values. Replace the missing values by the most frequent value in each column. Fit a decision tree classifier and observe the accuracy. Fit a random forest classifier and observe the accuracy..

Case Study 4: Principal component analysis using scikit learn.

Load the digits dataset from sklearn and write a helper function to plot the image. Fit a logistic regression model and observe the accuracy.

Using scikit learn perform a PCA transformation such that the transformed dataset can explain 95% of the variance in the original dataset. Compare it with a model and also comment on the accuracy. Compute the confusion matrix and count the number of instances that has gone wrong. For each of the wrong sample, plot the digit along with predicted and original label.

Case Study 5: Read the datafile “letterCG.data” and set all the numerical attributes as features. Split the data in to train and test sets.

Fit a sequence of AdaBoostClassifier with varying number of weak learners ranging from 1 to 16, keeping the max_depth as 1. Plot the accuracy on test set against the number of weak learners, using decision tree classifier as the base classifier.

Project #1:

Industry: Social Media

Problem Statement: You as ML expert have to do analysis and modeling to predict the number of shares of an article given the input parameters.

Actions to be performed:

Load the corresponding dataset. Perform data wrangling, visualization of the data and detect the outliers, if any. Use the plotly library in Python to draw useful insights out of data. Perform regression modeling on the dataset as well as decision tree regressor to achieve your goal. Also, use scaling processes, PCA along with boosting techniques to optimize your model to the fullest.

Project #2:

Industry: FMCG

Problem Statement: You as an ML expert have to cluster the countries based on various sales data provided to you across years.

Actions to be performed:

You have to apply an unsupervised learning technique like K means or Hierarchical clustering so as to get the final solution. But before that, you have to bring the exports (in tons) of all countries down to the same scale across years. Plus, as this solution needs to be repeatable you will have to do PCA so as to get the principal components which explain the max variance.