Parallel Implementation of Holt-Winters Algorithm for Big Data using MapReduce Programming Model

Arputhamary, B.; Arockiam, L.

doi:10.17577/ijertv5is020557

Cited by 2 publications

(4 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The final output is created by combining the results of all prediction. The following diagram illustrates the detailed steps of proposed algorithm in both of map and reduces stages [12].…”

Section: Prediction Methodologymentioning

confidence: 99%

Performance Improved Holt-Winter�s (PIHW) Prediction Algorithm for Big Data Environment

Arputhamary¹,

Arockiam²

2016

Int. Jour. Intel. Elec. Sys.

View full text Add to dashboard Cite

Prediction plays an important role everywhere particularly in business, technology and many others. It helps organizations to take timely decisions, to improve profits and to reduce lost sales. Recent years have witnessed an enormous development in the area of cloud computing and big data, which brings up challenges in decision making process. As the size of the dataset becomes extremely big, the process of extracting useful information by analysing these data has also become tedious. Today data are generated in an unprecedented manner, prediction plays major role in utilizing these data. Time Series based prediction models take great part in handling Big Data such as online sales data, weather data etc. In this paper a methodology for prediction is introduced and the model is evaluated by applying various time series models with time series data which is seasonal and non-stationary. From the analysis it is proved that Holt-Winter's model performs better in seasonal and non-stationary time series data. The Holt-Winters (HW) methods estimate three smoothing parameters, associated with level, trend and seasonal factors. The seasonal variation can be of either an additive or multiplicative form. Also in this paper, Performance Improved Holt-Winters (PIHW) prediction algorithm is proposed and the results demonstrate that a considerable reduction in forecast error (Mean Square Error) can be achieved in the proposed model compared to Holt-Winters (HW) model.

show abstract

“…The final output is created by combining the results of all prediction. The following diagram illustrates the detailed steps of proposed algorithm in both of map and reduces stages [12].…”

Section: Prediction Methodologymentioning

confidence: 99%

Performance Improved Holt-Winter�s (PIHW) Prediction Algorithm for Big Data Environment

Arputhamary¹,

Arockiam²

2016

Int. Jour. Intel. Elec. Sys.

View full text Add to dashboard Cite

show abstract

“…In Time Series analysis, forecasting plays an important role in the area of statistics, econometrics, quantitative finance, seismology, geophysics, weather, demand and sales forecasting. It is mainly used for signal detection and estimation in the context of signal processing, control and communication engineering [13]. In the context of data mining, pattern recognition and machine learning Time Series analysis can be used for clustering, classification and query by content, irregularity detection as well as forecasting.…”

Section: Aim and Objectivesmentioning

confidence: 99%

“…where  , β and γ are the smoothing parameters. 14The following section describes the experimentations of these forecasting models using CBS samples with R environment [12]- [14]. ISSN: 2278-9359 (Volume-6, Issue-8) V. EXPERIMENTAL SUMMARY The programming language R is used to build the Time Series models such as ARIMA, Holt-Winters Additive and Multiplicative with the aim of forecasting the bike counts for a certain set of season under various weather conditions.…”

Section: Arputhamary Et Al International Journal Of Emerging Researmentioning

confidence: 99%

“…Big Data Analytics (BDA) applications empower data scientists, predictive modellers, statisticians and other analytics professionals to analyze massive volumes of data [15]. These data are structured transaction data and also other forms of data such as semi-structured and un-structured that are often left idle by conventional Business Intelligence (BI) and other analytic programs [14]. The internet click stream data, web server logs, social media content, text from customer emails and survey responses, mobile-phone call-detail records and machine data captured by sensor devices that are connected to the Internet Of Things (IOT) are some of the examples for Big Data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Pragmatic Study on Time Series Models for Big Data

Arputhamary

Arockiam

2018

IJERMT

View full text Add to dashboard Cite

Recent years have witnessed the growth of Big Data, particularly Time Series data which initiates major research interest in Time Series analysis and forecasting future values. It finds interest in many applications such as business, stock market and exchange, weather forecasting, electricity demand, cost and usage of products and in any kind of place that has specific seasonal or trendy changes over time. The forecasting of Time Series data provides the organization with useful information that is necessary for making important decisions. In this paper, a detailed study is performed to find the total number of bike users with respect to the season and weather on Capital Bike Sharing System (CBS) dataset. The study covers the Auto Regressive Integrated Moving Average (ARIMA), Holt-Winters Additive and Multiplicative forecasting models to analyse the seasonal and trendy fluctuations of the given dataset to improve performance and accuracy.

show abstract

Parallel Implementation of Holt-Winters Algorithm for Big Data using MapReduce Programming Model

Cited by 2 publications

References 6 publications

Performance Improved Holt-Winter�s (PIHW) Prediction Algorithm for Big Data Environment

Performance Improved Holt-Winter�s (PIHW) Prediction Algorithm for Big Data Environment

A Pragmatic Study on Time Series Models for Big Data

Contact Info

Product

Resources

About