Interdisciplinary Studies for Better Learning Outcomes

Earlier, engineering was all about computer science, IT, civil and electrical, but now companies focus more on expertise such as machine learning and evolutionary computation, writes Shakti Sharma.

Hardest problems of the computer world are being solved by intelligent systems. From our mobile phones to clock, users want everything to be smart. To act smart, one needs to be intelligent, so humans are making intelligent systems. These intelligent systems are not always remodelled by using the structure of human brain. Today’s engineers also use intelligence of several other living species, such as ants, bees, fishes and monkeys. Initially, major research was focused on Artificial Neural Network (where the system tries to work on a neuron-based model like human brain) or genetic algorithms (by following rules of evolution). Evolutionary algorithm soon became popular as multi-agent system providing a better chance to find global optimisation.

Several multi-agent evolutionary algorithms were proposed over the last few decades. However, Ant Colony Optimisation (ACO) and Particle Swarm optimisation (PSO) became more popular than others. These algorithms are based on the intelligent acts of species such as ants, where they use a chemical compound pheromone to find the shortest path between their food and nest, and birds and fishes use a formation during their travel and they do not collide with each other. Now, humans are modelling that smartness and adding it into artificial systems. Currently, multiple companies are working on these models such as, Eurobios and Antoptima. Vehicle routing problem, scheduling problem, and computer vision are some of the major fields of application of these algorithms. In the race for providing more mechanisms, people started working on modelling insect of multiple species. However, researchers now feel that there should be enough evidence of benefits for developing a new biologically inspired algorithm. Since India is full of diversity, it provides huge scope of learning from social behaviour of different insects and animals, and with a proper research focus, one can use that learning for developing new algorithms to create smart systems. Course curriculum in engineering colleges should adapt to this requirement and should provide enough scope for incorporating research quotient right from the bachelor’s level. There was a time when engineering was all about computer science, information technology, civil and electrical, but now companies focus more on expertise such as machine learning and evolutionary computation. We are becoming increasingly dependent on delivery-based marketing and businesses. To cover more customers and provide efficient services, there is a huge requirement of resource optimisation techniques.

A very few universities provide evolutionary computation in their syllabus at bachelor’s level and some of them just provide the theoretical knowledge. One who learns algorithm should be able to use that algorithm on a bot. Hence, a B.Tech graduate should also be able to learn fundamentals of electrical engineering. Institutes should have proper focus on minor subjects, where the student can take a minor from another department while covering the required credit courses of their own subject. Proper interaction between faculties and students is essential for efficient delivery of knowledge and expertise.

(The author is assistant professor, Bennett University)