Today we are going to take about Machine Learning 101. I have rich experience in machine learning, deep learning, computer vision and artificial intelligence.

What is machine learning and its applications?
So let’s first understand the difference between AI, ML and deep learning. So artificial intelligence refers to incorporating human intelligence to machines.
Broadly this can be classified into two categories:
open-ended and close-ended.

So open-ended would mean that you can ask any question to an AI system and it would be able to interpret and reply back.
and Close-ended AI systems are built basically for a specific domain or a vertical.
Machine learning Machine learning refers to empowering computer systems to learn the pattern of trends in the data.
Like the terminology says, it’s all about making a machine learn the data so that when a new data is given to the machine it is able to predict the outcome with a certain probability. Machine learning can be broadly classified into supervised and unsupervised techniques. Supervised machine learning techniques would require obtaining data, whereas unsupervised can automatically find patterns from the data without the need of any training data. Deep learning So deep learning algorithms are inspired by the information processing patterns in the human brains. How are information patterns between different neurons in a similar fashion, artificial neural networks have been built with different layers where different weights get computed in between them in order to analyze the input it receives for a determined output. And all of this requires something called a foundational data science.

And so as you can see, deep learning is a sub set of machine learning which in turn is a sub set of artificial intelligence. Applications of machine learning Machine learning algorithms can help you recommend the right set of items when you shop on Amazon or Flipkart. Machine learning can also help you to reach your destination on time whenever you use your Google Maps with real time data feed. Machine learning can also help you to get the desired food when you order on your food delivery partners like Swiggy, Zomato etc. Journey in analytics So not only individuals but different organizations also need to journey in the field of data science. So they need to eventually mature and have different capabilities developed overtime. Both individuals as well as organizations will need to start from something, a phrase called as descriptive analytics. Descriptive analytics is all about telling what has happened in the past ruling on patterns and trends from the data. As they mature they need to move towards something called predictive analytics, like the terminology says it’s all about predicting what can happen in the future based on past trends or patterns. The next phrase is called prescriptive analytics, like how a doctor would diagnose his patient understand what are the diseases and predict what can happen if untreated.

Similarly, even a data scientist needs to prescribe to their clients when solving business problems by predicting what can happen in the future by analyzing the past data. Career options Different individuals have different levels of verticals in the data science, so they need to focus on reporting to descriptive analytics towards prescriptive analytics or ML & AI. Reporting and descriptive analytics have similar set of rules whereas predictive analytics and machine learning again have similar set of rules.

The skills required for recording would be like data manipulations, data management, data exploration, management information systems, they need to create reports, automate reports. In the descriptive analytics space the skills will be required like segmentation, customer profiling, portfolio analysis, trend analysis, forecasting. Different roles that are available in the market across different product bases at these companies are like management information systems analyst, data analyst, strategy analyst, cost analyst. In the predictive analytics space the skills required are like ability to build probabilistic models, classification models, regression trees, Bayesian statistics.

In the ML & AI area the skills you will be required to develop are around building neural networks, convolutional neural networks, recurrent neural networks, the ability to build LSTMs, geospatial models, fuzzy logic, inductive logic programming. So these are the most advanced skills that usually you need to develop to be able to prosper in the ML & AI space. And the roles that are available in predictive analytics and ML & AI space are like data scientist, statistical analyst, advanced analytics – team lead, market or global research analyst, analytics manager, vice president or director. And it will be both productive service based companies and few of them are for individual contributor unless you are a team manager. As a team manager you’ll need to mentor your team, share knowledge within the team, grow and mentor the team, build project road map and et cetera. But, as an individual contributor you will have to work on the project hands-on, right from the indulator doing proper analysis on the data, building models, validating and deploying each of the models.


Please enter your comment!
Please enter your name here