In the current era of artificial intelligence (AI) revolution, driven by technology breakthroughs like OpenAI’s ChatGPT, you may often hear the terms data science, machine learning (ML) and AI used interchangeably. “As these terms relate to career, skills and education, students must understand their difference and the process of evolution of the respective fields,” said Prof. Sankar Kumar Pal, National Science Chair, Government of India, and president of the Indian Statistical Institute. He was delivering a keynote address titled “Machine Intelligence and Data Science: Why and How?” at the Academy of Technology in Hooghly on the occasion of Engineers’ Day. Prof. Pal, a computer scientist, explained how the AI, ML and data mining fields all evolved from “pattern recognition”.
Pattern recognition is the process of discovering similarities within small problems to solve larger, more complicated problems. The technique is crucial in intelligent computer systems and proves useful in many application domains. Retailers use it to audit products using image recognition, financial analysts identify insights to understand and predict market trends and the police use it to sort through mounds of data and find patterns in various crimes.
“The technique of image processing emerged in the 1970s,” he said. This utilises pattern recognition and often a specific classification scheme to explore how to recognise image patterns. It also helps analyse videos to identify people, detect objects and enable autonomous driving. Post-2000s, pattern recognition became integral to extracting information and patterns from enormous data sets that emerged in the field of bioinformatics, such as the human genome project.
Today many experts consider “pattern recognition” an obsolete term and prefer “hot” topics, such as AI and deep learning, which teaches computers to process data in a way that is inspired by the human brain. “But that’s a mistake,” warned Prof. Pal “If one fails to understand the history and evolution of these fields, learning will be meaningless.” He mentioned how several new engineering colleges are indiscriminately introducing subjects like computer science and electronics (CSE) tagged with specialisations like AI and data science, without delving into the basics of pattern recognition.
Despite the current hype around AI, there is no machine that can compete with the human brain and its creative spark. Souvick Chatterjee, senior team lead at Mathworks, a company that develops mathematical computing software, focused on the importance of experiential learning for budding engineers. According to him, the idea of learning through experience, or “experiential learning”, has been around since 350BC when Aristotle mentioned “for the things we have to learn before we can do them, we learn by doing them”.
“Instead of reading about theories and concepts, you get to apply them in practical situations,” Chatterjee explained. This could be through projects, experiments, virtual experiments or even competitions. For example, building a robot, designing a bridge or creating a mobile app are all forms of experiential learning.
Prof. Anindita Banerjee, chairman trustee of the Academy of Technology, stressed that the spirit of engineering education is in experiential learning and innovation. She said, “This institute was founded by an engineer, the late Prof. Jagannath Banerjee. His aim was to nurture engineers among first generation learners and turn them into world class innovators.”
