Machine Learning

From translation apps to autonomous vehicles, all powers with Machine Learning. It offers a way to solve problems and answer complex questions. It is basically a process of training a piece of software called an algorithm or model, to make useful predictions from data. This article discusses the categories of machine learning problems, and terminologies used in the field of machine learning.

Types of machine learning problems

There are various ways to classify machine learning problems. Here, we discuss the most obvious ones.

 1. On basis of the nature of the learning “signal” or “feedback” available to a learning system

• Supervised learning: The model or algorithm is presented with example inputs and their desired outputs and then finds patterns and connections between the input and the output. The goal is to learn a general rule that maps inputs to outputs. The training process continues until the model achieves the desired level of accuracy on the training data. Some real-life examples are:
  • Image Classification: You train with images/labels. Then in the future, you give a new image expecting that the computer will recognize the new object.
  • Market Prediction/Regression: You train the computer with historical market data and ask the computer to predict the new price in the future.
• Unsupervised learning: No labels are given to the learning algorithm, leaving it on its own to find structure in its input. It is used for clustering populations in different groups. Unsupervised learning can be a goal in itself (discovering hidden patterns in data).
  • Clustering: You ask the computer to separate similar data into clusters, this is essential in research and science.
  • High-Dimension Visualization: Use the computer to help us visualize high-dimension data.
  • Generative Models: After a model captures the probability distribution of your input data, it will be able to generate more data. This can be very useful to make your classifier more robust.

A simple diagram that clears the concept of supervised and unsupervised learning is shown below:
 
As you can see clearly, the data in supervised learning is labeled, whereas data in unsupervised learning is unlabelled.

• Semi-supervised learning: Problems where you have a large amount of input data and only some of the data is labeled, are called semi-supervised learning problems. These problems sit in between both supervised and unsupervised learning. For example, a photo archive where only some of the images are labeled, (e.g. dog, cat, person) and the majority are unlabeled.
• Reinforcement learning: A computer program interacts with a dynamic environment in which it must perform a certain goal (such as driving a vehicle or playing a game against an opponent). The program is provided feedback in terms of rewards and punishments as it navigates its problem space.

 

2. Two most common use cases of Supervised learning are: 

• Classification: Inputs are divided into two or more classes, and the learner must produce a model that assigns unseen inputs to one or more (multi-label classification) of these classes and predicts whether or not something belongs to a particular class. This is typically tackled in a supervised way.  Classification models can be categorized in two groups: Binary classification and Multiclass Classification. Spam filtering is an example of binary classification, where the inputs are email (or other) messages and the classes are “spam” and “not spam”.
• Regression: It is also a supervised learning problem, that predicts a numeric value and outputs are continuous rather than discrete. For example, predicting stock prices using historical data.

An example of classification and regression on two different datasets is shown below: 

3. Most common Unsupervised learning are:

• Clustering: Here, a set of inputs is to be divided into groups. Unlike in classification, the groups are not known beforehand, making this typically an unsupervised task. As you can see in the example below, the given dataset points have been divided into groups identifiable by the colors red, green, and blue.
• Density estimation: The task is to find the distribution of inputs in some space.
• Dimensionality reduction: It simplifies inputs by mapping them into a lower-dimensional space. Topic modeling is a related problem, where a program is given a list of human language documents and is tasked to find out which documents cover similar topics.

On the basis of these machine learning tasks/problems, we have a number of algorithms that are used to accomplish these tasks. Some commonly used machine learning algorithms are Linear Regression, Logistic Regression, Decision Tree, SVM(Support vector machines), Naive Bayes, KNN(K nearest neighbors), K-Means, Random Forest, etc. Note: All these algorithms will be covered in upcoming articles.

Terminologies of Machine Learning

• Model A model is a specific representation learned from data by applying some machine learning algorithm. A model is also called a hypothesis.
• Feature A feature is an individual measurable property of our data. A set of numeric features can be conveniently described by a feature vector. Feature vectors are fed as input to the model. For example, in order to predict a fruit, there may be features like color, smell, taste, etc. Note: Choosing informative, discriminating and independent features is a crucial step for effective algorithms. We generally employ a feature extractor to extract the relevant features from the raw data.
• Target (Label) A target variable or label is the value to be predicted by our model. For the fruit example discussed in the features section, the label with each set of input would be the name of the fruit like apple, orange, banana, etc.
• Training The idea is to give a set of inputs(features) and its expected outputs(labels), so after training, we will have a model (hypothesis) that will then map new data to one of the categories trained on.
• Prediction Once our model is ready, it can be fed a set of inputs to which it will provide a predicted output(label). But make sure if the machine performs well on unseen data, then only we can say the machine performs well.

The figure shown below clears the above concepts: 

Here are the steps to get started with machine learning:

1. Define the Problem: Identify the problem you want to solve and determine if machine learning can be used to solve it.
2. Collect Data: Gather and clean the data that you will use to train your model. The quality of your model will depend on the quality of your data.
3. Explore the Data: Use data visualization and statistical methods to understand the structure and relationships within your data.
4. Pre-process the Data: Prepare the data for modeling by normalizing, transforming, and cleaning it as necessary.
5. Split the Data: Divide the data into training and test datasets to validate your model.
6. Choose a Model: Select a machine learning model that is appropriate for your problem and the data you have collected.
7. Train the Model: Use the training data to train the model, adjusting its parameters to fit the data as accurately as possible.
8. Evaluate the Model: Use the test data to evaluate the performance of the model and determine its accuracy.
9. Fine-tune the Model: Based on the results of the evaluation, fine-tune the model by adjusting its parameters and repeating the training process until the desired level of accuracy is achieved.
10. Deploy the Model: Integrate the model into your application or system, making it available for use by others.
11. Monitor the Model: Continuously monitor the performance of the model to ensure that it continues to provide accurate results over time.

Example :

Here is a simple machine-learning example in Python that demonstrates how to train a model to predict the species of iris flowers based on their sepal and petal measurements:

• Python

# Load the necessary libraries import pandas as pd from sklearn.model_selection import train_test_split from sklearn.svm import SVC    # Load the iris dataset df = pd.read_csv('iris.csv')    # Split the data into features and labels X = df[['sepal_length', 'sepal_width', 'petal_length', 'petal_width']] y = df['species']    # Split the data into training and testing sets X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)    # Create an SVM model and train it model = SVC() model.fit(X_train, y_train)    # Evaluate the model on the test data accuracy = model.score(X_test, y_test)    print('Test accuracy:', accuracy)

output:

Test accuracy: 0.9666666666666667
 

 Related Articles:

• Demystifying Machine Learning
• Machine Learning Applications

References:

• https://en.wikipedia.org/wiki/Machine_learning
• https://leonardoaraujosantos.gitbooks.io/artificial-intelligence/
• http://machinelearningmastery.com/data-terminology-in-machine-learning/

Introduction :

1. Getting Started with Machine Learning
2. An Introduction to Machine Learning
3. What is Machine Learning ?
4. Introduction to Data in Machine Learning
5. Demystifying Machine Learning
6. ML – Applications
7. Best Python libraries for Machine Learning
8. Artificial Intelligence | An Introduction
9. Machine Learning and Artificial Intelligence
10. Difference between Machine learning and Artificial Intelligence
11. Agents in Artificial Intelligence
12. 10 Basic Machine Learning Interview Questions

Data and It’s Processing:

1. Introduction to Data in Machine Learning
2. Understanding Data Processing
3. Python | Create Test DataSets using Sklearn
4. Python | Generate test datasets for Machine learning
5. Python | Data Preprocessing in Python
6. Data Cleaning
7. Feature Scaling – Part 1
8. Feature Scaling – Part 2
9. Python | Label Encoding of datasets
10. Python | One Hot Encoding of datasets
11. Handling Imbalanced Data with SMOTE and Near Miss Algorithm in Python
12. Dummy variable trap in Regression Models

Supervised learning :

1. Getting started with Classification
2. Basic Concept of Classification
3. Types of Regression Techniques
4. Classification vs Regression
5. ML | Types of Learning – Supervised Learning
6. Multiclass classification using scikit-learn
7. Gradient Descent :
   • Gradient Descent algorithm and its variants
   • Stochastic Gradient Descent (SGD)
   • Mini-Batch Gradient Descent with Python
   • Optimization techniques for Gradient Descent
   • Introduction to Momentum-based Gradient Optimizer
8. Linear Regression :
   • Introduction to Linear Regression
   • Gradient Descent in Linear Regression
   • Mathematical explanation for Linear Regression working
   • Normal Equation in Linear Regression
   • Linear Regression (Python Implementation)
   • Simple Linear-Regression using R
   • Univariate Linear Regression in Python
   • Multiple Linear Regression using Python
   • Multiple Linear Regression using R
   • Locally weighted Linear Regression
   • Generalized Linear Models
   • Python | Linear Regression using sklearn
   • Linear Regression Using Tensorflow
   • A Practical approach to Simple Linear Regression using R
   • Linear Regression using PyTorch
   • Pyspark | Linear regression using Apache MLlib
     
   • ML | Boston Housing Kaggle Challenge with Linear Regression
9. Python | Implementation of Polynomial Regression
10. Softmax Regression using TensorFlow
11. Logistic Regression :
    • Understanding Logistic Regression
    • Why Logistic Regression in Classification ?
    • Logistic Regression using Python
    • Cost function in Logistic Regression
    • Logistic Regression using Tensorflow
12. Naive Bayes Classifiers
13. Support Vector:
    • Support Vector Machines(SVMs) in Python
    • SVM Hyperparameter Tuning using GridSearchCV
    • Support Vector Machines(SVMs) in R
    • Using SVM to perform classification on a non-linear dataset
14. Decision Tree:
    • Decision Tree
    • Decision Tree Regression using sklearn
    • Decision Tree Introduction with example
    • Decision tree implementation using Python
    • Decision Tree in Software Engineering
15. Random Forest:
    • Random Forest Regression in Python
    • Ensemble Classifier
    • Voting Classifier using Sklearn
      
    • Bagging classifier

Unsupervised learning :

1. ML | Types of Learning – Unsupervised Learning
2. Supervised and Unsupervised learning
3. Clustering in Machine Learning
4. Different Types of Clustering Algorithm
5. K means Clustering – Introduction
6. Elbow Method for optimal value of k in KMeans
7. Random Initialization Trap in K-Means
8. ML | K-means++ Algorithm
9. Analysis of test data using K-Means Clustering in Python
10. Mini Batch K-means clustering algorithm
11. Mean-Shift Clustering
12. DBSCAN – Density based clustering
13. Implementing DBSCAN algorithm using Sklearn
14. Fuzzy Clustering
15. Spectral Clustering
16. OPTICS Clustering
17. OPTICS Clustering Implementing using Sklearn
18. Hierarchical clustering (Agglomerative and Divisive clustering)
19. Implementing Agglomerative Clustering using Sklearn
20. Gaussian Mixture Model

Reinforcement Learning:

1. Reinforcement learning
2. Reinforcement Learning Algorithm : Python Implementation using Q-learning
3. Introduction to Thompson Sampling
4. Genetic Algorithm for Reinforcement Learning
5. SARSA Reinforcement Learning
6. Q-Learning in Python

Dimensionality Reduction :

1. Introduction to Dimensionality Reduction
2. Introduction to Kernel PCA
3. Principal Component Analysis(PCA)
4. Principal Component Analysis with Python
5. Low-Rank Approximations
6. Overview of Linear Discriminant Analysis (LDA)
7. Mathematical Explanation of Linear Discriminant Analysis (LDA)
8. Generalized Discriminant Analysis (GDA)
9. Independent Component Analysis
10. Feature Mapping
11. Extra Tree Classifier for Feature Selection
12. Chi-Square Test for Feature Selection – Mathematical Explanation
13. ML | T-distributed Stochastic Neighbor Embedding (t-SNE) Algorithm
14. Python | How and where to apply Feature Scaling?
15. Parameters for Feature Selection
16. Underfitting and Overfitting in Machine Learning

Natural Language Processing :

1. Introduction to Natural Language Processing
2. Text Preprocessing in Python | Set – 1
3. Text Preprocessing in Python | Set 2
4. Removing stop words with NLTK in Python
5. Tokenize text using NLTK in python
6. How tokenizing text, sentence, words works
7. Introduction to Stemming
8. Stemming words with NLTK
9. Lemmatization with NLTK
10. Lemmatization with TextBlob
11. How to get synonyms/antonyms from NLTK WordNet in Python?

Neural Networks :

1. Introduction to Artificial Neutral Networks | Set 1
2. Introduction to Artificial Neural Network | Set 2
3. Introduction to ANN (Artificial Neural Networks) | Set 3 (Hybrid Systems)
4. Introduction to ANN | Set 4 (Network Architectures)
5. Activation functions
6. Implementing Artificial Neural Network training process in Python
7. A single neuron neural network in Python
8. Convolutional Neural Networks
   • Introduction to Convolution Neural Network
   • Introduction to Pooling Layer
   • Introduction to Padding
   • Types of padding in convolution layer
   • Applying Convolutional Neural Network on mnist dataset
9. Recurrent Neural Networks
   • Introduction to Recurrent Neural Network
   • Recurrent Neural Networks Explanation
   • seq2seq model
   • Introduction to Long Short Term Memory
   • Long Short Term Memory Networks Explanation
   • Gated Recurrent Unit Networks(GAN)
   • Text Generation using Gated Recurrent Unit Networks
10. GANs – Generative Adversarial Network
    • Introduction to Generative Adversarial Network
    • Generative Adversarial Networks (GANs)
    • Use Cases of Generative Adversarial Networks
    • Building a Generative Adversarial Network using Keras
    • Modal Collapse in GANs
11. Introduction to Deep Q-Learning
12. Implementing Deep Q-Learning using Tensorflow

ML – Deployment :

1. Deploy your Machine Learning web app (Streamlit) on Heroku
2. Deploy a Machine Learning Model using Streamlit Library
3. Deploy Machine Learning Model using Flask
4. Python – Create UIs for prototyping Machine Learning model with Gradio
5. How to Prepare Data Before Deploying a Machine Learning Model?
6. Deploying ML Models as API using FastAPI
   
7. Deploying Scrapy spider on ScrapingHub

ML – Applications :

1. Rainfall prediction using Linear regression
2. Identifying handwritten digits using Logistic Regression in PyTorch
3. Kaggle Breast Cancer Wisconsin Diagnosis using Logistic Regression
4. Python | Implementation of Movie Recommender System
5. Support Vector Machine to recognize facial features in C++
6. Decision Trees – Fake (Counterfeit) Coin Puzzle (12 Coin Puzzle)
7. Credit Card Fraud Detection
8. NLP analysis of Restaurant reviews
9. Applying Multinomial Naive Bayes to NLP Problems
10. Image compression using K-means clustering
11. Deep learning | Image Caption Generation using the Avengers EndGames Characters
12. How Does Google Use Machine Learning?
13. How Does NASA Use Machine Learning?
14. 5 Mind-Blowing Ways Facebook Uses Machine Learning
    
15. Targeted Advertising using Machine Learning
16. How Machine Learning Is Used by Famous Companies?

Misc :

1. Pattern Recognition | Introduction
2. Calculate Efficiency Of Binary Classifier
3. Logistic Regression v/s Decision Tree Classification
4. R vs Python in Datascience
5. Explanation of Fundamental Functions involved in A3C algorithm
6. Differential Privacy and Deep Learning
7. Artificial intelligence vs Machine Learning vs Deep Learning
8. Introduction to Multi-Task Learning(MTL) for Deep Learning
9. Top 10 Algorithms every Machine Learning Engineer should know
10. Azure Virtual Machine for Machine Learning
11. 30 minutes to machine learning
12. What is AutoML in Machine Learning?
13. Confusion Matrix in Machine Learning

Prerequisites to learn machine learning

• Knowledge of Linear equations, graphs of functions, statistics, Linear Algebra, Probability, Calculus etc.
• Any programming language knowledge like Python, C++, R are recommended.