ORIGINAL ARTICLE Solving 0–1 Knapsack Problem using Cohort Intelligence Algorithm Anand J. Kulkarni • Hinna Shabir Received: 26 December 2013 / Accepted: 28 May 2014 Ó Springer-Verlag Berlin Heidelberg 2014 Abstract An emerging technique, inspired from the nat- ural and social tendency of individuals to learn from each other referred to as Cohort Intelligence (CI) is presented. Learning here refers to a cohort candidate’s effort to self supervise its own behavior and further adapt to the behavior of the other candidate which it intends to follow. This makes every candidate improve/evolve its behavior and eventually the entire cohort behavior. This ability of the approach is tested by solving an NP-hard combinatorial problem such as Knapsack Problem (KP). Several cases of the 0–1 KP are solved. The effect of various parameters on the solution quality has been discussed.The advantages and limitations of the CI methodology are also discussed. Keywords Cohort Intelligence  Self Supervised Learning  Knapsack Problem  Combinatorial Optimization 1 Introduction The nature-/bio-inspired optimization techniques such as genetic algorithm (GA), particle swarm optimization (PSO), ant colony optimization (ACO), simulated annealing (SA), tabu search, etc., have become popular due to their simplicity to implement and working based on rules. The GA is population based which is evolved using the operators such as selection, crossover, mutation, etc. The performance of GA is governed by the quality of the population being evaluated and may often reach very close to the global optimal solution and necessitates local improvement techniques to incorporate into it [1, 2]. The paradigm of Swarm Intelligence (SI) is a decentralized self organizing optimization approach inspired from social behavior of living organisms such as insects, fishes, etc. which can communicate with one another either directly or indirectly. The techniques such as Particle Swarm Opti- mization (PSO) is inspired from the social behavior of bird flocking and school of fish searching for food [3]. The ACO works on the ants’ social behavior of foraging food following a shortest path [4]. Similar to ACO, the Bee Algorithm (BA) also works on the social behavior of honey bees finding the food; however, the bee colony tends to optimize the use of number of members involved in par- ticular predecided tasks [5]. The Cuckoo Search (CS) method exhibits slow convergence and marginally low accuracy. In addition, the Improved Cuckoo Search (ICS) uses adaptive step size to adjust search range, and genetic mutation operation to jump out of local optima [6]. Gen- erally, the swarm techniques are computationally intensive. All of these techniques are basically unconstrained opti- mization methods and their performance can be signifi- cantly affected when applied to constrained as well as combinatorial problems. An emerging Artificial Intelligence (AI) technique referred to as Cohort Intelligence (CI) was proposed in [7]. It is inspired from the self-supervised learning behavior of the candidates in a cohort. The cohort here refers to a group of candidates interacting and competing with one another A. J. Kulkarni (&) Odette School of Business, University of Windsor, 401 Sunset Avenue, Windsor, ON N9B 3P4, Canada e-mail: kulk0003@ntu.edu.sg; ajkulkarni@oatresearch.org; kulk0003@uwindsor.ca A. J. Kulkarni  H. Shabir Optimization and Agent Technology (OAT) Research Lab, Maharashtra Institute of Technology, 124 Paud Road, Kothrud, Pune 411038, India e-mail: me.hinna@gmail.com 123 Int. J. Mach. Learn. & Cyber. DOI 10.1007/s13042-014-0272-y