Authors :
Akarsh Prabhu K., Anukarsh G. Prasad , N. R. Prashanth , S.K.Padma.
Volume/Issue :
Volume 2 - 2017, Issue 6 - June
Google Scholar :
https://goo.gl/P2dRNm
Scribd :
https://goo.gl/mq5hnM
Thomson Reuters ResearcherID :
https://goo.gl/3bkzwv
Abstract :
Recent Trends in decision making have made use of various prediction models and algorithms to ensure accurate decisions. The prediction models are based on historic data aggregations. Statistical analysers train the models and portray results as predictions using need specific algorithms. The standout models use adaptive techniques to blend irregular variations in the data into the model; usage of weights during model design institutes the same. This paper conducts a thorough comparative study on different prediction models. It presents the results of time series prediction models and conclusively shows implementation of the linear regression modelling for Common Entrance Test results of Karnataka Education Authority. The model is built to forecast the future occurrences and rank colleges based on the predictions and compares the results of the three fore mentioned modelling techniques for the application. It incorporates outlier detection techniques for data pre processing. Further the data is optimised by assigning specific weights to particular data series based on trends. The paper introduces a breakthrough in the genre of decision support for career planning.
Keywords :
Linear Regression Model, Time-series, Rank Predictor, Holt Winters, ARIMA, R Programming.
Recent Trends in decision making have made use of various prediction models and algorithms to ensure accurate decisions. The prediction models are based on historic data aggregations. Statistical analysers train the models and portray results as predictions using need specific algorithms. The standout models use adaptive techniques to blend irregular variations in the data into the model; usage of weights during model design institutes the same. This paper conducts a thorough comparative study on different prediction models. It presents the results of time series prediction models and conclusively shows implementation of the linear regression modelling for Common Entrance Test results of Karnataka Education Authority. The model is built to forecast the future occurrences and rank colleges based on the predictions and compares the results of the three fore mentioned modelling techniques for the application. It incorporates outlier detection techniques for data pre processing. Further the data is optimised by assigning specific weights to particular data series based on trends. The paper introduces a breakthrough in the genre of decision support for career planning.
Keywords :
Linear Regression Model, Time-series, Rank Predictor, Holt Winters, ARIMA, R Programming.
