Machine learning-How it works?

What is machine learning?

Machine learning is a subfield of AI which focuses on the construction and study of systems that can learn from datasets. Here we will train the machine to learn from the previous datasets and make decisions and predictions based on experience. To get an accurate prediction we have to train the machine with immense and accurate dataset. There are various implementation of machine learning such as Tesla's self driving car, apple siri, sophia AI robot and many more.

Here the datasets which is to be fed into the machine will have one column to predict which is the result column. Training datasets is implemented to build up a model. Testing datasets is used to validate the model.

How does machine learning works?

One of the approaches is using labeled or unlabeled training datasets to produce the model. New input data is introduced to the machine learning algorithm and it makes predictions based on the model. The prediction is evaluated for accuracy.

If the accuracy is acceptable the machine learning algorithm is deployed.

Now if the accuracy is not acceptable the machine learning algorithm is strained again (i.e) we provide feedback to train the model and ask it to predict until it learns.

Life cycle of machine learning:

• Understand the business problem.
• Data collection.
• Exploratory data analysis and data cleaning.
• Machine learning algorithm.
• Predict your model accuracy.

Machine learning types:

• supervised learning.
• unsupervised learning.
• reinforcement learning.

Supervised machine learning:

This is used for classification problems or regression problems, where you map the input variables X to an output variable Y. Here each and every instance of training data has different input variables and an expected output variable called as labels or target class. So that the machine learns with those training examples and approximate the mapping function that is, when ever it have a new input data it predicts the output.

Unsupervised machine learning:

Here since training data has no labels it work on its own to discovers the pattern and groups the data according to the similarities in features. Unsupervised learning is further divided into clustering and association rule mining. Clustering is a process of grouping similar data items together. Association is a process of finding dependency of one data item on another item and map them accordingly which is more profitable.

Reinforcement learning:

Here there is no training samples, we will create an agent to do a particular task which learns by itself through interacting with environment. Performs some actions in environment and receives rewards based on those actions.

About srinila's view

we are here to explain about the booming technical field Artificial intellingence and the fields such as data science, machine learning, deep learning etc. which concerns with AI

Related Post