publications

2020

1. Fair Algorithms for Hierarchical Agglomerative Clustering

   Chhabra, Anshuman, Vashishth, Vidushi, and Mohapatra, Prasant

   2020

   Abs

   Hierarchical Agglomerative Clustering (HAC) algorithms are extensively utilized in modern data science, and seek to partition the dataset into clusters while generating a hierarchical relationship between the data samples. HAC algorithms are employed in many applications, such as biology, natural language processing, and recommender systems. Thus, it is imperative to ensure that these algorithms are fair – even if the dataset contains biases against certain protected groups, the cluster outputs generated should not discriminate against samples from any of these groups. However, recent work in clustering fairness has mostly focused on center-based clustering algorithms, such as k-median and k-means clustering. In this paper, we propose fair algorithms for performing HAC that enforce fairness constraints 1) irrespective of the distance linkage criteria used, 2) generalize to any natural measures of clustering fairness for HAC, 3) work for multiple protected groups, and 4) have competitive running times to vanilla HAC. Through extensive experiments on multiple real-world UCI datasets, we show that our proposed algorithm finds fairer clusterings compared to vanilla HAC as well as other state-of-the-art fair clustering approaches.

2. RLProph: a dynamic programming based reinforcement learning approach for optimal routing in opportunistic IoT networks

   Sharma, Deepak Kumar, Rodrigues, Joel J. P. C., Vashishth, Vidushi, Khanna, Anirudh, and Chhabra, Anshuman

   Wireless Networks 2020

   Abs

   Routing in Opportunistic Internet of Things networks (OppIoTs) is a challenging task because of intermittent connectivity between devices and the lack of a fixed path between the source and destination of messages. Recently, machine learning (ML) and reinforcement learning (RL) have been used with great success to automate processes in a number of different problem domains. In this paper, we seek to fully automate the OppIoT routing process by using the Policy Iteration algorithm to maximize the possibility of message delivery. Moreover, we model the OppIoT environment as a Markov decision process (MDP) replete with states, actions, rewards, and transition probabilities. The proposed routing protocol, RLProph, is able to optimize the routing process via the optimal policy obtained by solving the MDP using Policy Iteration. Through extensive simulations, we show that RLProph outperforms a number of ML-based and context-aware routing protocols on a multitude of performance criteria.

2019

1. GMMR: A Gaussian mixture model based unsupervised machine learning approach for optimal routing in opportunistic IoT networks

   Vashishth, Vidushi, Chhabra, Anshuman, and Sharma, Deepak Kumar

   Computer Communications 2019

   Abs

   Opportunistic IoT (OppIoT) network is a subclass of Internet of Things network, in which connections between the source and destination devices are intermittent. This infrequent connectivity is due to lack of network infrastructure and random mobility models followed by devices. These attributes of the network make routing in OppIoT, an increasingly complex problem. Moreover OppIoT shares its unique network characteristics with another class of networks called Opportunistic Networks (OppNets). This commonality enables the same routing designs to be applicable to both OppNets and OppIoT. Increased research interest in Machine Learning (ML) has led to its successful application in routing solutions for OppNets through protocols like KNNR and MLPROPH. In this paper we pursue utilizing ML to automate routing decisions in OppIoT. To this end we use Gaussian Mixture Models, an ML based soft clustering mechanism, to develop the proposed routing protocol called GMMR. The design of GMMR is such that it combines the advantages of both context-aware and context-free routing protocols. We compare the performance of GMMR with that of KNNR, HBPR, MLPROPH, and PROPHET using simulations run on Opportunistic Network Environment (ONE) simulator. The performance criteria for this comparison includes delivery probability, network overhead ratio, average hop count and number of messages dropped. We will show through the results of the simulations, that GMMR outperforms all of the aforementioned routing protocols in terms of every performance parameter.

2. A Machine Learning Approach Using Classifier Cascades for Optimal Routing in Opportunistic Internet of Things Networks

   Vashishth, Vidushi, Chhabra, Anshuman, and Sharma, Deepak Kumar

   In 2019 16th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON) 2019

2018

1. A fuzzy logic and game theory based adaptive approach for securing opportunistic networks against black hole attacks

   Chhabra, Anshuman, Vashishth, Vidushi, and Sharma, Deepak Kumar

   International Journal of Communication Systems 2018

   Abs

   Summary Opportunistic networks (OppNets) are networks that can be used in situations of sparse network connectivity. Strategies for implementing secure transmission of messages in OppNets offer complex and distinctive problems that are wholly different from traditional networks using TCP/IP protocols for data transmission. In this paper, a security protocol named FuzzyPT is proposed which combats black hole attacks in OppNets. The defense mechanism is based on the information extracted from messages in the buffer, potential threat (PT) messages, and fuzzy logic. Fuzziness aids the system in being adaptive by modeling the character of nodes as either benign or malicious. It provides the network with the ability to judge the character of nodes based on relationships between different parameters instead of basing this decision on fixed conditions. FuzzyPT is observed to reduce erroneous decision-making and decrease the number of false positives and false negatives in the system. An evolutionary game theoretic model is applied, which analyzes the decision-making ability of relays in choosing varied strategies while forwarding messages. The evolutionary stable state is computed for the protocol, which verifies the proposition that a majority of the relays will cooperate by forwarding messages. The efficiency of the protocol is judged on various grounds such as delivery probability, network overhead ratio, packets dropped, and transmission latency. It is observed that the proposed game theoretic protocol outperforms an existing acknowledge and evolutionary game theory–based security protocol in terms of all the aforementioned performance characteristics.

2017

1. A game theory based secure model against Black hole attacks in Opportunistic Networks

   Chhabra, Anshuman, Vashishth, Vidushi, and Sharma, Deepak Kumar

   In 2017 51st Annual Conference on Information Sciences and Systems (CISS) 2017