Volume 10 Number 1 June 2020


Is Deep Autoencoder a Better Forecasting Tool? - Case of Employee Attrition
Kalyan Sankar Sengupta, T. K. Mithuna,T. Aashika

Abstract-Human Resource (HR) manager is responsible for extracting the useful information or pattern from the past, present and future employees of an organization in order to retain the best employees. The behavior and intelligence of an employee can be analyzed in many ways using statistical models, mathematical models and business intelligence techniques. In fact, all the organizations or industries are having good database about the details of employees. For any organization, the role of employee is inevitable and the most of the companies are using innovative ideas to retain the recruited people. Nowadays, with the existence of Data Science and prediction techniques, this task can be automatically done, which allows the managers of the companies to obtain the information they require from the employees in a much faster and efficient way than it was obtained in the past when the task was done manually by the human resources department. These results in a significant decrease of the costs associated with employee attrition, turnover or churn, maximizing the revenue of the company. Many researchers have explored the data mining techniques such as decision tree algorithm, Back Propagation Neural Network, and Random Forest method. In this paper, a novel method based on machine learning has been proposed to predict the employee attrition using deep autoencoder. Deep Autoencoder consists of two halves, one representing the encoding half and the other representing the decoding half. The model has been validated is compared with the results of Na�ve Bayes, Support Vector Machine (SVM), Back Propagation Neural Network (BPN) and Random Forest.


A Study of Stock Market Price using Sentimental Analysis on Banking Sector
Kalyan Sengupta, Aman Abhishek Kisku, Srujana R, Prakhar Singhal

Abstract- Interpreting stock market has been an area of interest for some time now. Social media now a days represents the sentiment of public and opinion about ongoing events. Stock market interpretation using twitter on basis of sentiment of the public is an attractive research field. Sentiment analysis includes the process of extracting useful information from various books and social media platforms like Twitter, Facebook to understand about the emotions of the people about an event. The sentiment of the collected data can be analysed using various languages and tools like Python, R, KNIME, Orange. For any investor it is important to invest in a stock which would give decent return and that would perform consistently withstanding the market tension. Nowadays, with the availability of Data Science and interpretation techniques, this task can be automatically done, which will help the investors to make better investment decision. Many researchers have used Python to interpret the sentiment of the data. In this paper, we have used Orange to obtain the polarity scores which will used to interpret the sentiment of the data.


Modelling and Solving Differential Equations using Neural Networks: A Study
R. Devipriya, S. Selvi

Abstract- - Generally, Neural Networks (NN) are considered as a hierarchical models that can be used to learn patterns or knowledge from data with complicated nature or distribution. These NNs are also used as universal function approximators. Therefore, NNs can be applied to solve the mathematical problems, as numerical analysis tool. This paper discusses applications of neural networks in modelling and finding solution of various differential equations.


A Study on the Role of Big Data Analytics in Organizations
P. Anbumani,K. Selvaraj

Abstract- Massive facts is everywhere and there's almost an urgent need to accumulate and hold anything records is being generated, for the worry of missing out on something crucial there's a massive quantity of records floating around. What we do with it's far all that matters right now that is why big records Analytics is within the frontiers of IT. Massive information analytics has end up crucial as it aids in improving enterprise, choice makings and providing the biggest side over the competitors. This applies for groups in addition to experts within the Analytics domain. For experts, who're professional in massive records Analytics, there may be an ocean of possibilities out there.


Applications of Biometrics Personal Authentication (Mobile & Computers)
G. Ramachandran, S. Kannan, G. Murali, P.M . Murali,T. Sheela,T.Muthumanickam

Abstract- Nowadays, we are talking more and more about insecurity in various sectors in addition because the computer techniques to be implemented to counter this trend: access control to computers, e-commerce, banking, etc. There are two traditional ways of identifying a private. the primary method may be a knowledge-based method. it's supported the knowledge of an individual�s information like the PIN code to permit him/her to activate a portable. The second method is predicated on the possession of token. It will be a bit of identification, a key, a badge, etc. These two methods of identification will be employed in a complementary thanks to obtain increased security like in bank cards. However, they each have their weaknesses. Within the first case, the password will be forgotten or guessed by a 3rd party. Within the second case, the badge (or ID or key) could also be lost or stolen. Biometric features are an alternate solution to the 2 previous identification modes. The advantage of using the biometric features is that they're all universal, measurable, unique, and permanent. The interest of applications using biometrics will be summed up in two classes: to facilitate the way of life and to avoid scam.