
Instructor
Manaranjan Pradhan

Category
Neural Network and AI

Course Fees
Quotation on request Rs.
Course name : Neural_Network_and_AI_5_Days
BUILDING AI SYSTEM USING PYTHON
GOOGLE’S SELFDRIVING CARS AND ROBOTS GET A LOT OF PRESS, BUT THE COMPANY’S REAL FUTURE IS IN MACHINE LEARNING, THE TECHNOLOGY THAT ENABLES COMPUTERS TO GET SMARTER AND MORE PERSONAL.
– ERIC SCHMIDT (GOOGLE CHAIRMAN)
This course is intended to give a holistic understanding on statistical & machine learning and its application using Python. The workshop will cover
 An introduction to business analytics
 An introduction to Python for data analysis
 An introduction to supervised machine learning algorithms
 An introduction to unsupervised machine learning algorithms
 Understanding the core of machine learning – Gradient Descent Algorithm
 Understanding of various sampling strategies and its efficacy in learning process
 An introduction to ensemble methods for handling imbalanced data
 Introduction to text analytics using Python
 Handson using the Python code on the real life dataset
OBJECTIVE
We are living in an era where computing moved from mainframes to personal computers to cloud. And while it happened, we started generating humongous amount of data. However the multifolds increase in computing power also brought in advancement in application of algorithms which can be used to get insights from huge amount of data being generated. In this course, you will learn to nuances of building machine learning models on real life datasets. We’ll introduce you to Python platform and some of the machine learning algorithms which may be used for building an AI system.
At the end of the course you will develop a clear understanding of the evolution of Artificial Intelligence system and how neural network and other machine learning algorithms forms the integral part of it.
WHO SHOULD ATTEND
Irrespective of type of industry (retail, ecommerce, manufacturing, real estate & construction, telecom, hospitality, banking, healthcare, IT, supply chain &logistic, etc.); data forms the crux of decision making. This course is designed hone up analytical skills and business acumen of midlevel and senior level corporate professionals trying to understand the nuances of data science and help them the machine learning techniques an efficient way to generate insights for customers which in turn optimizes the bottom line of organizations.
HARDWARE AND SOFTWARE
 Participants should bring their laptop (preferably Windows 7 or higher/ Mac OS installed).
 Operating System (any of the following):
 Mac OS X with XQuartz
 Windows (Version XP or later) is required.
 Minimum 8 GB RAM on the system is advisable. 16GB RAM is preferable.
INSTALLATIONS:
 For Windows, go to https://docs.anaconda.com/anaconda/install/windows.html
 For MacOS, go to https://docs.anaconda.com/anaconda/install/macos
 For Linux, go to https://docs.anaconda.com/anaconda/install/linux
More about anaconda can be found at https://docs.anaconda.com. Participants are expected to resolve any installation issues of the software prior to the commencement of the session.
PREREQUISITE & COURSE DELIVERABLE
 Participants should have basic programming skills. Participants are expected to spend time with the code set as a home assignment to leverage the classroom training hours to the fullest.
 High speed internet connection will be provided at the training venue.
 Deliverable: Python code and dataset. Soft copy of the content being covered (PDF file)
COURSE OUTLINE
Day 1: Understanding Anaconda Framework platform and other useful packages in Python
Session 1&2–Introduction to Business Analytics
 What is Business Analytics
 Why is it needed and how industries are adopting it
 Different components of analytics
 Applications of analytics in different domains
 Statistical learning vs. Machine learning
 Artificial Intelligence, Machine Leaning and Deep Learning
 What is Data Science and skills of a data scientist
 Different types of machine learning algorithms–Supervised, Unsupervised and Reinforcement learning
Session 3, 4 & 5–Introduction to Anaconda and Python
 Overview of Anaconda framework
 Python – Variables, objects, loops, conditions, function.
 Python Data structures – lists, tuples, dictionaries, sets
 Introduction to Numpy – ndarrays, ndarrays indexing, ndarrays datatypes and operations, statistical sorting and set operation
 Introduction to Pandas – Data ingestion, descriptive statistics, visualization, frequent data operations, merging dataframes, parsing timestamps
 Introduction to visualization – Matplotlib
 Introduction to H2O package, AutoML, Driverless AI
Day 2: Understanding advanced algorithms and its implementation using Python
Session 1, 2 & 3–Lab 1: Linear Regression
 Introduction to simple and multiple linear regression
 Regression diagnostic–Rsquared, ttest, Ftest, error terms distribution, heteroscedasticity, identifying multicollinearity and handling, AIC  model selection strategy
 Common task framework for model evaluation – training and test set.
 Case study using regression techniques and handson using Python code for regression
Session 4&5–Lab 2: Logistic Regression
 Introduction to logistic regression
 Logistic regression diagnostic: Wald statistics, Hosmer Lemeshow test, Classification Matrix, Sensitivity, Specificity, ROC Curve, precision, recall, F1score
 Strategy to find the optimal cutoff
 Bias and variance in the model, Bias vs. variance tradeoff
 Case study using logistic regression techniques and handson using Python code for regression
Day 3: Understanding supervised learning and gradient descent algorithm
Session 1 & 2 –Lab 2: Logistic Regression
 Introduction to logistic regression
 Logistic regression diagnostic: Classification Matrix, Sensitivity, Specificity, ROC Curve, precision, recall, F1score
 Strategy to find the optimal cutoff , cost of misclassification, F Score strategy
 Bias and variance in the model, Bias vs. variance tradeoff
 Case study using logistic regression techniques and handson using Python code for regression
Session 3 & 4 –Lab 3: Introduction to Gradient Descent
 Hypothesis formulation for linear regression
 Deriving the cost function for linear regression
 Cost function– Intuition for linear regression with one parameter and two parameters
 Gradient descent algorithm–application in linear regression
 Hypothesis formulation for logistic regression
 Deriving the cost function for logistic regression
 Cost function– Intuition for logistic regression
 Gradient descent algorithm–application in logistic regression
 Handson using Python code to implement gradient descent for regression and logistic regression
Session 5–Lab 5: Decision Trees
 Decision tree – Classification and regression trees (CART), Gini Index, Entropy
 Decision tree – Chisquare automatic interaction detection (CHAID)
 Case study using decision tree techniques
 Handson using Python code
Day 4: Understanding unsupervised learning, ensemble methods and text analytics
Session 1–Lab 5: Clustering and Segmentation
 Supervised and Unsupervised learning
 Clustering–Hierarchical, K means
 Clustering diagnostic–Dendrogram , Calinski and Harabasz index, Silhouette width
 Case study using hierarchical clustering and K–means clustering tree techniques
 Handson using Python code for Hierarchical and K–means cluster
Session 2, 3 & 4–Lab 6: Other Machine learning models (Ensemble Methods)
 What is Machine learning
 Different sampling strategies–Bootstrapping, Up–Sample, Down–Sample, Synthetic Sample, Cross–Validation Data
 Introduction to Bagging–Random Forest, Other Bagging algorithms
 Introduction to Boosting– Adaptive boosting, Other Boosting algorithms
 Case study of an imbalanced data and application of sampling strategies & ensemble methods
 Handson using Python code on an imbalanced data
Session 5–Lab 7: Introduction to Natural Language Processing
 Reading in unstructured dataset
 Data preprocessing – Stop words, stemming, lemmatization
 Tokenization and bag of words
 Plotting word frequency
 Sentiment analysis using Python NLTK
Day 5: Understanding unsupervised learning, ensemble methods and text analytics
Session 1&2: Introduction to Neural Network
 History of Neural network and its analogy to human intelligence
 Basic perceptron model and its working
 Neural Model with a logistic unit
 Neural network architecture – for binary and multiclass classification
 Hyper parameter Tuning
 Neural Network using H2O package
Session 3–Lab 7: Neural Network architecture and gradient descent
 Feedforward neural network, vectorized implementation  Feedforward neural network
 Neural Network – Feature Learning
 Neural Network – Cost function
 Gradient computation for Neural Network – Back propagation algorithm
 Implementation of Neural Network from through gradient descent
Session 2–Lab 6: Multivariate Gaussian Model for Anomaly Detection
 What is a fraud/anomaly in finance data
 Multivariate Gaussian Model for anomaly detection
Hands on using python code to implement anomaly detection
COURSE SCHEDULE
Day 1: Understanding Anaconda Framework platform and other useful packages in Python
This day will be primarily cover introduction to business analytics, introduction to Anaconda and Python
Introduction to Business Analytics 
1 
9 AM 
10:15 AM 
Introduction to Business Analytics 
2 
10:30 AM 
11:15 AM 
Introduction to Anaconda and Python platform 
3 
12:00 PM 
1:15 PM 
Introduction to Anaconda and Python platform…cont. 
4 
2:15 PM 
3:30 PM 
Introduction to Anaconda and Python platform…cont. 
5 
3:45 PM 
5:00 PM 
Day 2: Understanding the concepts of Regression and Logistic Regression
Day is primarily devoted to concept building on supervised learning and handson using Python code for the same
Lab 1: Multiple Linear Regression 
1 
9 AM 
10:15 AM 
Lab 1: Multiple Linear Regression…cont. 
2 
10:30 AM 
11: 45 AM 
Lab 1: Multiple Linear Regression…cont. 
3 
12:00 PM 
1:15 PM 
Lab 2: Logistic Regression 
4 
2:15 PM 
3:30 PM 
Lab 2: Logistic Regression…cont. 
5 
3:45 PM 
5:00 PM 
Day 3: Understanding gradient descent algorithm for Regression and Logistic Regression
Day will cover concept building on using gradient descent algorithm
Lab 2: Logistic Regression…cont. 
1 
9 AM 
10:15 AM 
Lab 2: Logistic Regression…cont. 
2 
10:30 AM 
11: 45 AM 
Lab 3: Introduction to Gradient descent 
3 
12:00 PM 
1:15 PM 
Lab 3: Introduction to Gradient descent…cont. 
4 
2:15 PM 
3:30 PM 
Lab 4: Decision Tree 
5 
3:45 PM 
5:00 PM 
Day 4: Understanding unsupervised learning, ensemble methods and text analytics
Day will cover concept building on unsupervised learning, sampling strategy and text analytics
Lab 5: Clustering and Segmentation 
1 
9 AM 
10:15 AM 
Lab 6: Other Machine leaning models 
2 
10:30 AM 
11: 45 AM 
Lab 6: Other Machine leaning models…cont. 
3 
12:00 PM 
1:15 PM 
Lab 6: Other Machine leaning models…cont. 
4 
2:15 PM 
3:30 PM 
Lab 7: Introduction to NLP 
5 
3:45 PM 
5:00 PM 
Day 5: Understanding unsupervised learning, ensemble methods and text analytics
Day will cover concept building on unsupervised learning, sampling strategy and text analytics
Lab 7: Introduction to Neural Network 
1 
9 AM 
10:15 AM 
Lab 7: Introduction to Neural Network…cont. 
2 
10:30 AM 
11: 45 AM 
Lab 7: Neural Network architecture and gradient descent 
3 
12:00 PM 
1:15 PM 
Lab 7: Neural Network architecture and gradient descent 
4 
2:15 PM 
3:30 PM 
Multivariate Gaussian Model for Anomaly Detection 
5 
3:45 PM 
5:00 PM 
John Doe says:
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna et sed aliqua. Ut enim ea commodo consequat...
ReplayJohn Doe says:
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna et sed aliqua. Ut enim ea commodo consequat...
ReplayJohn Doe says:
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna et sed aliqua. Ut enim ea commodo consequat...
ReplayJohn Doe says:
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna et sed aliqua. Ut enim ea commodo consequat...
ReplayJohn Doe says:
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna et sed aliqua. Ut enim ea commodo consequat...
Replay