Unsupervised Learning :
It’s a type of learning where we don’t give target to our model while training i.e. training model has only input parameter values. The model by itself has to find which way it can learn. Data-set in Figure A is mall data that contains information of its clients that subscribe to them. Once subscribed they are provided a membership card and so the mall has complete information about customer and his/her every purchase. Now using this data and unsupervised learning techniques, mall can easily group clients based on the parameters we are feeding in.
Training data we are feeding is –
- Unstructured data: May contain noisy(meaningless) data, missing values or unknown data
- Unlabeled data : Data only contains value for input parameters, there is no targeted value(output). It is easy to collect as compared to labelled one in Supervised approach.
Types of Unsupervised Learning :-
- Clustering: Broadly this technique is applied to group data based on different patterns, our machine model finds. For example in above figure we are not given output parameter value, so this technique will be used to group clients based on the input parameters provided by our data.
- Association: This technique is a rule based ML technique which finds out some very useful relations between parameters of a large data set. For e.g. shopping stores use algorithms based on this technique to find out relationship between sale of one product w.r.t to others sale based on customer behavior. Once trained well, such models can be used to increase their sales by planning different offers.
As the name suggests, its working lies between Supervised and Unsupervised techniques. We use these techniques when we are dealing with a data which is a little bit labelled and rest large portion of it is unlabeled. We can use unsupervised technique to predict labels and then feed these labels to supervised techniques. This technique is mostly applicable in case of image data-sets where usually all images are not labelled.
In this technique, model keeps on increasing its performance using a Reward Feedback to learn the behavior or pattern. These algorithms are specific to a particular problem e.g. Google Self Driving car, AlphaGo where a bot competes with human and even itself to getting better and better performer of Go Game. Each time we feed in data, they learn and add the data to its knowledge that is training data. So, more it learns the better it get trained and hence experienced.
- Agents observe input.
- Agent performs an action by making some decisions.
- After its performance, agent receives reward and accordingly reinforce and the model stores in state-action pair of information.
- Temporal Difference (TD)
- Deep Adversarial Networks
Hope you liked the article