In a Machine Leaning field, one of the majour tasks is classification (others are regression, clustering and recommendation). Today I will explain the concept of classification by using a daily example.
Classification is labeling or judging a output target into multiple categories based on some input information. This output should be infinite number of categories.
For example, based on the weight and height, judging the grade would be one simplest example. Of course this classification is assumed to have lots of misclassification. Also note that grades seemed to be numbers but in this case we regard them as categories.
Another example. based on recorded voice mails, if you would like to identify the gender of the speaker, it will be classfication.
A more complex example would be hand-written digit classification. If you send physical mIl, not email, the zipcode will be scanned in a machinery in USPS, and it will automatically classify the scanned image into 5 digit, which can be readable from other systems.
Input data can be numbers or categories, or mixture of them. Audio record and images of zipcode are made of numbers if you look at the data in a computer.
Having a baby, I am now building a classifier to identify the categories what the baby wants by screaming e.g., milk, changing a diaper, rubbing his back to burp.