Mining and Forecasting of Big Time-Series Data

  title={Mining and Forecasting of Big Time-Series Data},
  author={Yasushi Sakurai and Yasuko Matsubara and Christos Faloutsos},
  journal={2019 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)},
Given a large collection of time series, such as motion capture sensors and automobile trajectories, how can we efficiently and effectively find typical patterns? How can we statistically summarize all the sequences, and achieve a meaningful segmentation? What are the major tools for fore-casting and outlier detection? Time-series data analysis becomes of increasingly high importance, thanks to the decreasing cost of hardware and the increasing online processing abilities. The objective of our… 

Fast Mining and Forecasting of Co-evolving Epidemiological Data Streams

This paper proposes a new streaming algorithm, EPICAST, which is able to model, understand and forecast dynamical patterns in large co-evolving epidemiological data streams, and is based on a unified non-linear differential equation.

Regime Shifts in Streams: Real-time Forecasting of Co-evolving Time Sequences

REGIMECAST is designed as an adaptive non-linear dynamical system, which is inspired by the concept of "regime shifts" in natural dynamical systems and outperforms state-of-the-art competitors as regards accuracy and speed.

Feature-aware forecasting of large-scale time series data sets

CSAR, a technique forecasting a set of time series with only one model, and a feature-aware partitioning applying CSAR on subsets of similar time series provide accurate forecasts a hundred times faster than traditional techniques, preparing forecasting for the arising challenges of the IoT era.

Automatic Sequential Pattern Mining in Data Streams

This paper proposes a streaming algorithm, namely StreamScope, that is designed to find intuitive patterns efficiently from event streams evolving over time and can achieve great improvements in terms of computational time and memory space over its full batch method competitors.

Non-Linear Mining of Social Activities in Tensor Streams

A streaming method that is designed to capture basic trends and seasonality in tensor streams and extract temporal and multi-dimensional relationships between such dynamics, and outperforms the state-of-the-art algorithms for time series forecasting in terms of forecasting accuracy and computational time.

A Novel Hybrid Method for Time Series Forecasting Using Soft Computing Approach

  • A. SanghaniN. BhattN. Chauhan
  • Computer Science
    Proceedings of the International Conference on ISMAC in Computational Vision and Bio-Engineering 2018 (ISMAC-CVB)
  • 2019
This work has projected a hybrid model of ARIMA and SVM, which has demonstrated great outcomes in solving nonlinear regression estimation problems and to utilize the linear strength of ARimA.

Driving with Data: Modeling and Forecasting Vehicle Fleet Maintenance in Detroit

This work utilizes tensor decomposition techniques to discover and visualize unique temporal patterns in vehicle maintenance; applies differential sequence mining to demonstrate the existence of common and statistically unique maintenance sequences by vehicle make and model; and demonstrates an application of a predictive Long Short Term Memory (LSTM) neural network model to predict maintenance sequences.

StreamScope: Automatic Pattern Discovery over Data Streams

This paper proposes a streaming algorithm, namely StreamScope, that is designed to find intuitive patterns efficiently from event streams evolving over time and can achieve great improvements in terms of computational time and memory space over its full batch method competitors.

Analyzing Load Profiles of Electricity Consumption by a Time Series Data Mining Framework

The experimental results show that the dimension reduction method known as piecewise aggregate approximation can help to detect the state of the annealing furnace and detect abnormal patterns of the load profile of their electricity consumption.



Streaming Pattern Discovery in Multiple Time-Series

SPIRIT can incrementally capture correlations and discover trends, efficiently and effectively, and be used to immediately spot potential anomalies, to do efficient forecasting and to dramatically simplify further data processing.

Online data mining for co-evolving time sequences

This work develops a fast method to analyze co-evolving time sequences jointly to allow estimation/forecasting of missing/delayed/future values, quantitative data mining, and outlier detection, and adapts to changing correlations among time sequences.

Searching and Mining Trillions of Time Series Subsequences under Dynamic Time Warping

This work shows that by using a combination of four novel ideas the authors can search and mine truly massive time series for the first time, and shows that in large datasets they can exactly search under DTW much more quickly than the current state-of-the-art Euclidean distance search algorithms.

Fast and Exact Monitoring of Co-Evolving Data Streams

The experiments on 67GB of real data illustrate that Stream Scan does indeed detect the qualifying subsequence patterns correctly and that it can offer great improvements in speed (up to 479,000 times) over its competitors.

Fast mining and forecasting of complex time-stamped events

TriMine is introduced, which performs three-way mining for all three attributes, namely, URLs, users, and time, and consistently outperforms the best state-of-the-art existing methods in terms of accuracy and execution speed.

F4: large-scale automated forecasting using fractals

A fast, automated method to do non-linear forecasting, for both periodic as well as chaotic time series, using the technique of delay coordinate embedding and using the concept of `intrinsic dimensionality'.

F-Trail: Finding Patterns in Taxi Trajectories

A novel method, called F-Trail, is developed, which allows to find meaningful patterns and anomalies from a huge collection of taxi trajectories, and is effective, leading to novel discoveries, and surprising outliers.

Optimal multi-scale patterns in time series streams

This work introduces a method to discover optimal local patterns, which concisely describe the main trends in a time series and introduces a criterion to select the best window sizes, which most concisely capture the key oscillatory as well as aperiodic trends.

Stream Monitoring under the Time Warping Distance

A theoretical analysis is provided and it is proved that SPRING does not sacrifice accuracy, while it requires constant space and time per time-tick, and that it can offer dramatic improvements in speed over the naive implementation.

AutoPlait: automatic mining of co-evolving time sequences

The method has the following properties: effectiveness: it operates on large collections of time-series, and finds similar segment groups that agree with human intuition, and scalability: it is linear with the input size, and thus scales up very well.