Machine Learning Image

An Introduction to Machine Learning

MACHINE LEARNING

Humans learn from past experiences,  Machines follow the instructions given by humans but, what if humans can train the machines to learn from the past experiences (data) and can do act much faster, here comes the concept of Machine Learning.

What is Machine learning?

Machine learning is the field of study that gives computers the capability to learn without being explicitly programmed. Machine learning algorithms build a mathematical model based on the data, known as training data, in order to make predictions or decisions. Machine learning is not only about learning, but also about understanding and reasoning. Machine Learning is not programmed, it is taught with data. The foundation of effective machine learning is useful data that is Big data.

Machine Learning Image

ARTIFICIAL INTELLIGENCE VS MACHINE LEARNING:

Artificial Intelligence means machines can perform tasks in ways that are “intelligent”. They are not programmed to do a single repetitive task, they adapt to different situations. Machine Learning is a subpart or branch of Artificial intelligence. ML is an application of AI that provides the system with the ability to automatically learn and improve from experience. In ML we can generate a program by integrating input & output or only inputs of that program.

Machine Learning vs AI

TYPES OF MACHINE LEARNING:
  • Supervised Learning

  • Unsupervised Learning

SUPERVISED LEARNING:

In supervised learning, it requires input and output pairs. We have to feed input with associated outputs i.e., labeled dataset. If this task repeats several times with many data, the algorithm picks up the pattern between the inputs and outputs. Then if we feed the new input it will predict the output based on the pattern obtained from the previous data.

Apple as a input data and using algorithm to process the raw data and receive the output form machine

Let us take this as an example. Previously the machine doesn’t know anything about apples. Then you are training the machine with some data known as training data, like if the input data is red color and round in shape then it is an apple. The machine understands and labels the classification. Now if you give new input the machine recognizes it as an apple by the previous pattern.

Recommending new songs based on someone’s past music choices is an example of supervised learning. The model is training a classifier on pre-existing labels (genres of songs). This is what Netflix and Spotify do, they collect the songs/movies that you like already, evaluate the features based on your likes or dislikes and then recommend new movies &songs based on similar features.

ALGORITHMS USED IN SUPERVISED LEARNING:
  • Polynomial regression

  • Random forest

  • Linear regression

  • Logistic regression

  • Decision trees

  • K-nearest neighbors

  • Naive Bayes.

UNSUPERVISED LEARNING:

In unsupervised learning, the input is fed without associated output. On repeating the steps the algorithm clusters the input into groups. If a new input is fed the algorithm will predict, which cluster the new data belongs to.

Input Data and Model trained output

 

Let us take this example for unsupervised learning. Previously the machine doesn’t know anything about fruits. Then you are providing some set of data. Since the machine doesn’t have any previous experience, it separates data into groups based on similar features. If you provide new input, it groups the new data into the respective cluster.

Analyzing bank data for suspicious-looking transactions and flag fraud transactions is an example of unsupervised learning. In this case, the suspicious transactions are not defined, hence there are no labels of “fraud” and “not fraud”. The model tries to identify outliers by looking at anomalous transactions and flags them as ‘fraud’.

ALGORITHMS USED IN UNSUPERVISED LEARNING:
  • K-means clustering

  • Partial least squares

  • Fizzy means

  • Hidden Markov models

  • Hierarchical clustering

  • Principal component analysis

  • Self-organizing maps.

USES OF MACHINE LEARNING:

Machine learning has wide applications like Email spam filtering, Web search results, pattern and image recognition, Video recommendation, and so on. ML becoming a more widely accepted and adapted technology.

Its capabilities are widely applied and it is changing the business in dynamic ways.

Applications of machine learning

APPLICATIONS OF MACHINE LEARNING:

Most importantly Machine learning is being applied in the health care field to improve patient care and help avoid lapses that occur due to human error. It is more efficient to detect diseases such as early-stage cancer which often goes unnoticed at an early stage.

It is being employed by most social media networks to provide a more personalized and enjoyable experience.

PREREQUISITES TO BECOME AN MACHINE LEARNING ENGINEER:
  • Basic knowledge of programming and scripting languages.

  • Intermediate knowledge of statistics and probability

  • Knowledge of data structures and algorithms.

 

To Explore More about Emerging Technologies  – 

Leave a Comment