Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

"Time Series Analysis": models, code, and papers

Event-driven timeseries analysis and the comparison of public reactions on COVID-19

Apr 30, 2021
Md. Khayrul Bashar

The rapid spread of COVID-19 has already affected human lives throughout the globe. Governments of different countries have taken various measures, but how they affected people lives is not clear. In this study, a rule-based and a machine-learning based models are applied to answer the above question using public tweets from Japan, USA, UK, and Australia. Two polarity timeseries (meanPol and pnRatio) and two events, namely "lockdown or emergency (LED)" and "the economic support package (ESP)", are considered in this study. Statistical testing on the sub-series around LED and ESP events showed their positive impacts to the people of (UK and Australia) and (USA and UK), respectively unlike Japanese people that showed opposite effects. Manual validation with the relevant tweets showed an agreement with the statistical results. A case study with Japanese tweets using supervised logistic regression classifies tweets into heath-worry, economy-worry and other classes with 83.11% accuracy. Predicted tweets around events re-confirm the statistical outcomes.

* 15 pages, 10 figures, 6 tables, International Conference on Big Data and Application (BDAP 2021), for associated file, see 

  Access Paper or Ask Questions

Impact of novel aggregation methods for flexible, time-sensitive EHR prediction without variable selection or cleaning

Sep 17, 2019
Jacob Deasy, Ari Ercole, Pietro Liò

Dynamic assessment of patient status (e.g. by an automated, continuously updated assessment of outcome) in the Intensive Care Unit (ICU) is of paramount importance for early alerting, decision support and resource allocation. Extraction and cleaning of expert-selected clinical variables discards information and protracts collaborative efforts to introduce machine learning in medicine. We present improved aggregation methods for a flexible deep learning architecture which learns a joint representation of patient chart, lab and output events. Our models outperform recent deep learning models for patient mortality classification using ICU timeseries, by embedding and aggregating all events with no pre-processing or variable selection. Our model achieves a strong performance of AUROC 0.87 at 48 hours on the MIMIC-III dataset while using 13,233 unique un-preprocessed variables in an interpretable manner via hourly softmax aggregation. This demonstrates how our method can be easily combined with existing electronic health record systems for automated, dynamic patient risk analysis.

* 5 pages, 3 tables, 1 figure, preprint under review at the Machine Learning for Health workshop at NeurIPS 2019 

  Access Paper or Ask Questions

Stack Index Prediction Using Time-Series Analysis

Aug 18, 2021
Raja CSP Raman, Rohith Mahadevan, Divya Perumal, Vedha Sankar, Talha Abdur Rahman

The Prevalence of Community support and engagement for different domains in the tech industry has changed and evolved throughout the years. In this study, we aim to understand, analyze and predict the trends of technology in a scientific manner, having collected data on numerous topics and their growth throughout the years in the past decade. We apply machine learning models on collected data, to understand, analyze and forecast the trends in the advancement of different fields. We show that certain technical concepts such as python, machine learning, and Keras have an undisputed uptrend, finally concluding that the Stackindex model forecasts with high accuracy and can be a viable tool for forecasting different tech domains.

* 10 pages, 9 figures 

  Access Paper or Ask Questions

A Comparative Study on Forecasting of Retail Sales

Mar 14, 2022
Md Rashidul Hasan, Muntasir A Kabir, Rezoan A Shuvro, Pankaz Das

Predicting product sales of large retail companies is a challenging task considering volatile nature of trends, seasonalities, events as well as unknown factors such as market competitions, change in customer's preferences, or unforeseen events, e.g., COVID-19 outbreak. In this paper, we benchmark forecasting models on historical sales data from Walmart to predict their future sales. We provide a comprehensive theoretical overview and analysis of the state-of-the-art timeseries forecasting models. Then, we apply these models on the forecasting challenge dataset (M5 forecasting by Kaggle). Specifically, we use a traditional model, namely, ARIMA (Autoregressive Integrated Moving Average), and recently developed advanced models e.g., Prophet model developed by Facebook, light gradient boosting machine (LightGBM) model developed by Microsoft and benchmark their performances. Results suggest that ARIMA model outperforms the Facebook Prophet and LightGBM model while the LightGBM model achieves huge computational gain for the large dataset with negligible compromise in the prediction accuracy.

  Access Paper or Ask Questions

A multiscale spatiotemporal approach for smallholder irrigation detection

Feb 09, 2022
Terence Conlon, Christopher Small, Vijay Modi

In presenting an irrigation detection methodology that leverages multiscale satellite imagery of vegetation abundance, this paper introduces a process to supplement limited ground-collected labels and ensure classifier applicability in an area of interest. Spatiotemporal analysis of MODIS 250m Enhanced Vegetation Index (EVI) timeseries characterizes native vegetation phenologies at regional scale to provide the basis for a continuous phenology map that guides supplementary label collection over irrigated and non-irrigated agriculture. Subsequently, validated dry season greening and senescence cycles observed in 10m Sentinel-2 imagery are used to train a suite of classifiers for automated detection of potential smallholder irrigation. Strategies to improve model robustness are demonstrated, including a method of data augmentation that randomly shifts training samples; and an assessment of classifier types that produce the best performance in withheld target regions. The methodology is applied to detect smallholder irrigation in two states in the Ethiopian highlands, Tigray and Amhara. Results show that a transformer-based neural network architecture allows for the most robust prediction performance in withheld regions, followed closely by a CatBoost random forest model. Over withheld ground-collection survey labels, the transformer-based model achieves 96.7% accuracy over non-irrigated samples and 95.9% accuracy over irrigated samples. Over a larger set of samples independently collected via the introduced method of label supplementation, non-irrigated and irrigated labels are predicted with 98.3% and 95.5% accuracy, respectively. The detection model is then deployed over Tigray and Amhara, revealing crop rotation patterns and year-over-year irrigated area change. Predictions suggest that irrigated area in these two states has decreased by approximately 40% from 2020 to 2021.

  Access Paper or Ask Questions