Welcome to AI/ML/DL

This post will be the driving factor for all the Artificial Intelligence, Machine Learning & Deep Learning Content. 
Well here you can checkout the differences.

1. Part 0: Grooming Sessions
• Machine Learning Applied
2. Part 1: Data Preprocessing 
3. Part 2: Regression 
• Simple Linear Regression
• Multiple Linear Regression
• Polynomial Regression
• Support Vector Regression (SVR)
• Decision Tree Regression
• Random Forest Regression
• Evaluating Regression Models Performance
4. Part 3: Classification 
• Logistic Regression
• K-Nearest Neighbors (K-NN)
• Support Vector Machine (SVM)
• Kernel SVM
• Naive Bayes
• Decision Tree Classification
• Random Forest Classification
• Evaluating Classification Models Performance
5. Part 4: Clustering 
• K-Means Clustering
• Hierarchical Clustering
6. Part 5: Association Rule Learning 
• Apriori
• Eclat
7. Part 6: Reinforcement Learning 
• Upper Confidence Bound (UCB)
• Thompson Sampling
8. Part 7: Natural Language Processing 
9. Part 8: Deep Learning 
• Artificial Neural Networks
• Convolutional Neural Networks
10. Part 9: Dimensionality Reduction 
• Principal Component Analysis (PCA)
• Linear Discriminant Analysis (LDA)
• Kernel PCA
11. Part 10: Model Selection & Boosting 
• Model Selection
• XGBoost

Model Selection Technique - Link, Link, Link
Dataset available at - Link

Types of Machine Learning Algorithms

1. Supervised Learning 
1. Regression
2. Classification
2. Un-Supervised Learning
1. Clustering
2. Association Analysis
3. Reinforcement learning
4. Dimension reduction

Reference - Video Link

1). Regression:

• Goal - Predicting Numeric Values
• Used for predicting continuous values.
• E.g.
• Predicting Stock Price
• Determine Sales demand for next year
• Predict next 24 hrs rain
• Determine likelihood of medicine effectiveness of a patient 
• Predicting grades of students studying from 0-6 hrs

2). Classification: 

• Goal - Predict Category
• Binary - Predicting Stock Price High/Low, Yes/No, Dog/Cat, Pass/Fail, Male/Female, 
• Multiple - Sunny/Cloudy/Rainy/Windy, Risk - High/Medium/Low, Digits Entered - 0-9, Sentiment - Positive, Negative, Neutral
• When the data are being used to predict a categorical variable, supervised learning is also called classification.
• When there are only two labels, this is called Binary classification. 
• When there are more than two categories, the problems are called Multi-class classification.

3). Clustering: 

• Goal - Organize Similar Items Into Respective Groups.
• Grouping a set of data examples so that examples in one group (or one cluster) are more similar (according to some criteria) than those in other groups. 
• Often used to segment the whole dataset into several groups
• Analysis can be performed in each group to help users to find intrinsic patterns.
• E.g. 
• Customer Segmentation - Seniors,Adults, Teenagers, Kids
• Areas of similar topography - Desert, Grass, Water etc.

4). Association Analysis: 

• Goal - Capture association between items.
• E.g.
• Identify Items purchased together
• Identify Web pages visited together

5). Dimension reduction: Reducing the number of variables under consideration. In many applications, the raw data have very high dimensional features and some features are redundant or irrelevant to the task. Reducing the dimensionality helps to find the true, latent relationship.

6). Reinforcement learning:
Reinforcement learning analyzes and optimizes the behavior of an agent based on the feedback from the environment.  Machines try different scenarios to discover which actions yield the greatest reward, rather than being told which actions to take. Trial-and-error and delayed reward distinguishes reinforcement learning from other techniques.

Supervised learning
With supervised learning, you have an input variable that consists of past labeled training data and a desired output variable. E.g Age & Weight to derive Human Health. You use an algorithm to analyze the training/past/labelled data to learn the function that maps the input to the output. This inferred function maps new, unknown examples by generalizing from the training data to anticipate results in unseen situations.

Semi-supervised learning
The challenge with supervised learning is that labeling data can be expensive and time consuming. If labels are limited, you can use unlabeled examples to enhance supervised learning. Because the machine is not fully supervised in this case, we say the machine is semi-supervised. With semi-supervised learning, you use unlabeled examples with a small amount of labeled data to improve the learning accuracy.

Unsupervised learning
When performing unsupervised learning, the machine is presented with totally unlabeled data. Here we solve Clustering (Organize Similar Items Into Respective Group) kind of problems. It is asked to discover the intrinsic patterns that underlies the data, such as a Clustering structure, Low-dimensional manifold, or a Sparse tree and Graph.

Few Use Cases:

• Automate Invoice Reading thru OCR 
• Fraud detection (eg. If approval limit is $1000 and multiple approvals are happening $999.99)  
• While Creating Order, when a supplier is added, past performance of that supplier can be shown to the buyer (as contextual insight) to better equip the buyer with a sound reason of whether or not he needs to proceed with the same supplier
• Built Auto-Intelligence in Comment box
• While Creating Order, with the help of ML, we can know a particular user creates what type of orders mostly that user creates and auto-fill ORDER
• Reduce Mobile Fraud - Added Face Recognition While Creating Order 

Hope this helps!!

Arun Manglick