Multi Clustering Recommendation System for Fashion Retail

Fashion retail has a large and ever-increasing popularity and relevance, allowing customers to buy anytime finding the best offers and providing satisfactory experiences in the shops. Consequently, Customer Relationship Management solutions have been enhanced by means of several technologies to better understand the behaviour and requirements of customers, engaging and influencing them to improve their shopping experience, as well as increasing the retailers’ profitability. Current solutions on marketing provide a too general approach, pushing and suggesting on most cases, the popular or most purchased items, losing the focus on the customer centricity and personality. In this paper, a recommendation system for fashion retail shops is proposed, based on a multi clustering approach of items and users’ profiles in online and on physical stores. The proposed solution relies on mining techniques, allowing to predict the purchase behaviour of newly acquired customers, thus solving the cold start problems which is typical of the systems at the state of the art. The presented work has been developed in the context of Feedback project partially founded by Regione Toscana, and it has been conducted on real retail company Tessilform, Patrizia Pepe mark. The recommendation system has been validated in store, as well as online.

COVID-19 impact on Socio-economic and Health Interventions: A Gaps and Peaks analysis using Clustering Approach

A quantifiable model to describe the peaks and gaps during the several waves of COVID 19 is generated and applied to the progression of 120 countries. The number of waves encountered and how many more to be encountered is a question which is currently explored by all the scientific communities. In the same quest, an attempt has been made to quantitatively model the peaks and the gaps within them which have been encountered by 120 most affected countries from February 2020 to December 2021. These 120 countries were ranked based on the number of confirmed cases and deaths recorded during this period. This study further cluster these countries based on socio economic and health interventions to find an association with three dependent features of COVID 19 i.e. number of confirmed cases, deaths and death infectivity rate. The findings in this study suggests that, every wave had multiple peaks within them and as the number of peaks increased, predicting their growth rate or decline rate turns to be extremely difficult. However, considering the clusters which share the common features even with diverse countries, there is some possibility to predict what might be coming next. This study involves exhaustive analysis of reliable data which are available in open access and marks an important aspect to the COVID 19 research communities.

A Hybrid Classification Technique for Enhancing the Effectiveness of Intrusion Detection Systems Using Machine Learning

The objective of this paper is to propose and develop a hybrid intrusion detection system to handle series and non-series data by applying the two different concepts that are named clustering and autocorrelation function in a single architecture. There is a need to propose and build a system that can handle both types of data whether it is series or non-series. Therefore, the authors used two concepts to generate a robust approach to craft a hybrid intrusion detection system. The authors utilize an unsupervised clustering approach that is used to categorize the data based on domain similarity to handle non-series data and another approach is based on autocorrelation function to handle series data. The approach is consumed in single architecture where it carries data as input from both host-based intrusion detection systems and network-based intrusion detection systems. The result shows that the hybrid intrusion detection system is categorizing data based on the optimal number of clusters obtained through the elbow method in clustering.