MACHINE LEARNING & IT'S VARIOUS TECHNIQUES
MACHINE LEARNING
Machine learning (ML) is the study of the algorithms and statistical models which are used by systems in order to increase the performance on a specific task in a progressive manner .
It is a branch of the artificial intelligence which is based on the idea that systems can learn from data , they can identify patterns & can also make decisions with minimal human intervention .
EVOLUTION OF MACHINE LEARNING
Because of the new technological advancements , the ML of today has got a huge difference as compared to the past scenario . It was started with pattern recognition and at that time it was not much adaptive.
Researchers , who are interested in artificial intelligence , willing to see if computers can learn from the data and this iterative aspect of ML was important so as when the models are exposed to new data
they are able to independently adapt it . They learn from previous computations in order to produce reliable & repeatable decisions & results . It's the science that was not new , but that , which has got a momentum.
There are some wide applications of ML which we are familiar with :
- The heavily hyped _self driving google car -? an essence of machine learning .
- Online recommendations such as those from amazon and Netflix ? machine learning for everyday life .
- Knowing what customers are saying about you on Twitter ? ML combined with the linguistic creation.
- Fraud detection ? one of the obvious and important ML application of today's world .
WHAT'S REQUIRED TO CREATE GOOD MACHINE LEARNING SYSTEMS
The needs to create good machine learning systems are :
- Data preparation capabilities .
- Algorithms - basics and advanced
- Automation and iterative processes.
- Scalability.
- Modeling .
Some Facts:
- In ML , a target is called as a label.
- In statistics , a target is called a variable.
- A variable in statistics is called a feature in machine learning.
- A transformation in statistics is called feature creation in machine learning.
POPULAR MACHINE LEARNING TECHNIQUES
Some of the popular machine learning techniques are :
- Supervised learning
- Semi supervised learning
- Unsupervised learning
- Reinforcement learning
- SUPERVISED LEARNING :
These algorithms are trained using labeled examples , such as an input where the desired output is known.The learning algorithm receives a set of inputs along with the corresponding correct outputs , & algorithm learns by comparing its actual output to correct outputs to find the errors '
for eg; it can anticipate when credit card transactions are likely to be fraud or which an insurance customer is likely to file a claim .
2. SEMI SUPERVISED LEARNING :
It is used for the same applications as supervised learning . It used both i.e., the labeled and unlabeled data for training - typically a small amount of labeled data along with a large amount of unlabeled data . Semi supervised learning is used in the case when cost associated with the labeling is too high to allow a labeled training process .
3 .UNSUPERVISED LEARNING :
It is used against the data that has no historical labels . The system is not told the " right answer ". The algorithm needs to figure out , what is being shown . The goal is to explore the data & find some structure within. It works well on the transaction data.
4. REINFORCEMENT LEARNING :
With the reinforcement learning , the algorithm discovers through the trial and error methods i.e.; which action yields the best reward. This type of learning has got three components
- Agent ( a learner or a decision maker )
- Environment ( the things agent interacts with)
- Actions ( which the agent can do)
Nice work nikhil, good content
ReplyDeleteThank you brother
Delete