A comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMM
Time series prediction is one of the main areas of statistics and machine learning. In 2018 the two new algorithms higher order hidden Markov model and temporal convolutional network were proposed and emerged as challengers to the more traditional recurrent neural network and long-short term memory...
Main Author: | |
---|---|
Format: | Others |
Language: | English |
Published: |
Högskolan i Skövde, Institutionen för informationsteknologi
2019
|
Subjects: | |
Online Access: | http://urn.kb.se/resolve?urn=urn:nbn:se:his:diva-16974 |
id |
ndltd-UPSALLA1-oai-DiVA.org-his-16974 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-UPSALLA1-oai-DiVA.org-his-169742019-06-11T04:39:20ZA comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMMengAlmqvist, OlofHögskolan i Skövde, Institutionen för informationsteknologi2019machine learningdeep learningtime series forecastingtime series regressiondata sciencepredictioncrisp-dmkerasmarkov modelneural networkexploratory data analysismaskininlärningdjupinlärningtidsserieprediktiontidsserieprognosneurala nätverkmarkovmodellexplorativ dataanalysdataanalysEngineering and TechnologyTeknik och teknologierTime series prediction is one of the main areas of statistics and machine learning. In 2018 the two new algorithms higher order hidden Markov model and temporal convolutional network were proposed and emerged as challengers to the more traditional recurrent neural network and long-short term memory network as well as the autoregressive integrated moving average (ARIMA). In this study most major algorithms together with recent innovations for time series forecasting is trained and evaluated on two datasets from the theme park industry with the aim of predicting future number of visitors. To develop models, Python libraries Keras and Statsmodels were used. Results from this thesis show that the neural network models are slightly better than ARIMA and the hidden Markov model, and that the temporal convolutional network do not perform significantly better than the recurrent or long-short term memory networks although having the lowest prediction error on one of the datasets. Interestingly, the Markov model performed worse than all neural network models even when using no independent variables. Student thesisinfo:eu-repo/semantics/bachelorThesistexthttp://urn.kb.se/resolve?urn=urn:nbn:se:his:diva-16974application/pdfinfo:eu-repo/semantics/openAccess |
collection |
NDLTD |
language |
English |
format |
Others
|
sources |
NDLTD |
topic |
machine learning deep learning time series forecasting time series regression data science prediction crisp-dm keras markov model neural network exploratory data analysis maskininlärning djupinlärning tidsserieprediktion tidsserieprognos neurala nätverk markovmodell explorativ dataanalys dataanalys Engineering and Technology Teknik och teknologier |
spellingShingle |
machine learning deep learning time series forecasting time series regression data science prediction crisp-dm keras markov model neural network exploratory data analysis maskininlärning djupinlärning tidsserieprediktion tidsserieprognos neurala nätverk markovmodell explorativ dataanalys dataanalys Engineering and Technology Teknik och teknologier Almqvist, Olof A comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMM |
description |
Time series prediction is one of the main areas of statistics and machine learning. In 2018 the two new algorithms higher order hidden Markov model and temporal convolutional network were proposed and emerged as challengers to the more traditional recurrent neural network and long-short term memory network as well as the autoregressive integrated moving average (ARIMA). In this study most major algorithms together with recent innovations for time series forecasting is trained and evaluated on two datasets from the theme park industry with the aim of predicting future number of visitors. To develop models, Python libraries Keras and Statsmodels were used. Results from this thesis show that the neural network models are slightly better than ARIMA and the hidden Markov model, and that the temporal convolutional network do not perform significantly better than the recurrent or long-short term memory networks although having the lowest prediction error on one of the datasets. Interestingly, the Markov model performed worse than all neural network models even when using no independent variables. |
author |
Almqvist, Olof |
author_facet |
Almqvist, Olof |
author_sort |
Almqvist, Olof |
title |
A comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMM |
title_short |
A comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMM |
title_full |
A comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMM |
title_fullStr |
A comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMM |
title_full_unstemmed |
A comparative study between algorithms for time series forecasting on customer prediction : An investigation into the performance of ARIMA, RNN, LSTM, TCN and HMM |
title_sort |
comparative study between algorithms for time series forecasting on customer prediction : an investigation into the performance of arima, rnn, lstm, tcn and hmm |
publisher |
Högskolan i Skövde, Institutionen för informationsteknologi |
publishDate |
2019 |
url |
http://urn.kb.se/resolve?urn=urn:nbn:se:his:diva-16974 |
work_keys_str_mv |
AT almqvistolof acomparativestudybetweenalgorithmsfortimeseriesforecastingoncustomerpredictionaninvestigationintotheperformanceofarimarnnlstmtcnandhmm AT almqvistolof comparativestudybetweenalgorithmsfortimeseriesforecastingoncustomerpredictionaninvestigationintotheperformanceofarimarnnlstmtcnandhmm |
_version_ |
1719202515561480192 |