Page 160 - AI Ver 1.0 Class 10
P. 160

Unsupervised Learning
              An unsupervised learning approach works on an unlabeled dataset. This means that the data which is fed to the
              machine is random and there is no knowhow available about it to the trainer.
              These learning models are used to identify trend, pattern and relationship in the data which is fed into it. In this
              model the major features are identified by the machine, which helps the user in understanding the data. For
              example, in the data of 100 cat images, if you want to understand some pattern in the data, you would need to
              feed this data into the unsupervised learning model and train the machine. Once trained, the machine would
              identify patterns in the data. These patterns might already be known to the user, like colour or size, or something
              unusual about the cats.
              Unsupervised Learning can further be divided into:
                 • Clustering: It is a machine learning approach where the machine partitions the dataset into different clusters
                or categories based on machine generated algorithms. The data fed to such a model is usually unlabeled or
                random and thus the developer feeds in the data directly into the machine and instructs it to build its own
                algorithm. The machine then forms a pattern or cluster based on training data and groups those that follow the
                same pattern.























                 The best clustering is the one that minimizes the error.  Clustering works on discrete dataset. For example, if you
                have random data of flowers from your garden, since you are unable to find any meaningful pattern amongst
                them, you would feed their data into the clustering algorithm. The algorithm would then analyse the data and
                divide them into clusters according to their similarities based on the trends noticed. The clusters are then given
                as the output.

                 • Dimensionality Reduction: Humans can visualise any figure up to 3-Dimensions only but according to a
                lot of theories and algorithms, there are various entities which exist beyond 3-Dimensions. For example,
                in Natural language Processing, the words are considered to be N-Dimensional entities. Which means
                that we cannot visualise them as they exist beyond our visualisation ability. Hence, to make sense out of
                it, we need to reduce their dimensions which we do by using dimensionality reduction algorithm. As we
                reduce the dimension of an entity, the information which it contains starts getting distorted. For example,
                if we have a ball in our hand, it is 3-Dimensional right now. But if we click its picture, the data transforms
                to 2-D as an image which is a 2-Dimensional entity. Now, as soon as we reduce one dimension, at least
                50% of the information is lost as now we will not know about the back of the ball. Whether the ball was
                of the same colour at the back or not? Or was it just a hemisphere? If we reduce the dimensions further,
                more and more information will get lost. So, we use Dimensionality Reduction here which reduces the
                dimensions and makes it sensible data.



                        158   Touchpad Artificial Intelligence-X
   155   156   157   158   159   160   161   162   163   164   165