You will be surprised to know that artificial intelligence, one of the most sought after technologies of this era has been predicted to create a value of at least 13 trillion dollars by the year 2030. But what exactly is artificial intelligence.
Artificial intelligence can be described as the intelligence demonstrated by machines in contrast to natural intelligence displayed by humans and animals. Let’s understand this through an example.
Suppose there is a room inside which a human is sitting and who can translate a given language say for example English into some another language Chinese. Now if we replace that human with such a machine that can do the translation. Now we being the the user of such a translation machine, if we are unable to perceive difference between the the human translator and the machine translator, then the machine is said to be exhibiting intelligence.
The below given diagram illustrates the various subsets of of artificial intelligence.
First let us understand the basics of machine learning popularly known as ML
Machine learning can be described as the ability of computers to learn and complete tasks without being explicitly programmed
Machine learning or ML is a subset of artificial intelligence. In classical programming problems we used to code specific set of instructions and and then get results as per our requirement. But there are many real world problems in which we exactly do not know what could be the rules to set.
For example suppose we want to identify whether a fruit is an apple or an orange.Another example problem could be to identify whether an animal is a cat or a dog. In all of such problems we ourselves do not know how to convert such problems into a program. In all of such real world problems , machine learning comes to our help. The main idea is to create a learning model and then train it with sample of data available with us. As a model gets trained, its accuracy improves and it is capable of predicting . See diagram below for more clarity.
To accept the prediction or the forecast of our machine learning model, first we need to define our acceptance criteria. That is how much accuracy we want in the prediction of our model.
Machine learning is helpful in solving many real world problems.
- Early prediction of cancer
- Stock market prediction
- Finding suitable diagnosis
- Video playlist suggestions
- Designing personalized medication
- Speech recognition
- Image recognition
- And many more………..
Let us take another example from our city Muzaffarpur. Suppose someone living in Bihar wants to set up a real estate company. If he is capable of predicting houseprices in all the localities of Muzaffarpur, it will help his real estate company grow. For this he starts collecting data of houses from different localities of Muzaffarpur. (One sample data set can contain these information-
- Built Area of house
- Location in the city
- Nearby landmarks like Muzaffarpur Railway station,motijheel business area
- No of rooms
- No of floors
- Availability of amenities like lift, power backup
- Price of the house
In the terminology of Machine Learning, all of the above are known as features and the variable that we want to predict is called as prediction variable.
Now that person willing to open the real estate company will build one ML model and train it with all the data he has acquired. Once the model is trained, it will be able to predict price of a house by taking input features. So if his customers provide him with their requirements, he can easily quote them approximate cost of the house.
I hope that by following my article, you must have by now gained some idea about machine learning or artificial intelligence. Still if there is something about which you are unclear, let me know in the comments section.
Other places where you can get resources for machine learning and AI are