eda-bench Tasks

Implement anomaly detection algorithms for time-series data. Identify outliers and unusual patterns using statistical and machine learning methods.

anomaly detection, outliers, time-series, isolation forest, statistical methods, python, scikit-learn, pattern recognition

Created by Artur Dox

Compute ACF and PACF plots to identify AR/MA model orders and periodicity in time-series. Apply Box-Jenkins methodology with stationarity checks before analysis.

autocorrelation, ACF, PACF, partial autocorrelation, ARIMA, Box-Jenkins, AR order, MA order, periodicity, stationarity, ADF test, KPSS, statsmodels, python

Created by Artur Dox

Apply change point detection algorithms to identify structural breaks and regime changes in time-series data. Use statistical and machine learning approaches.

change point detection, CUSUM, PELT, time-series, structural breaks, python, ruptures, bayesian, regime change

Created by Artur Dox

Analyze channel correlation patterns in time-series data. Identify and quantify correlations between different data channels using statistical methods.

time-series, correlation, signal processing, python, pandas, numpy, statistical analysis

Created by Artur Dox

Compute cross-correlation functions between multiple time-series channels to identify lead-lag dependencies. Apply pre-whitening and significance testing for robust lag estimation.

cross-correlation, lag analysis, CCF, cross-correlogram, pre-whitening, lead-lag, multi-channel, time-series, python, statsmodels, numpy, synchronization

Created by Artur Dox

Resample and align multi-rate time-series data onto a common time grid using interpolation and anti-aliased decimation. Validate temporal alignment accuracy across channels.

resampling, interpolation, time alignment, upsampling, downsampling, anti-aliasing, decimation, PCHIP, cubic spline, multi-rate, time-series, pandas, scipy, python, synchronization

Created by Artur Dox

Detect signal envelopes using the Hilbert transform and peak interpolation. Extract upper and lower envelopes for amplitude modulation and energy analysis.

envelope detection, Hilbert transform, analytic signal, instantaneous amplitude, peak picking, amplitude modulation, RMS, time-series, scipy, python, spline interpolation

Created by Artur Dox

Extract comprehensive features from time-series data for machine learning. Implement statistical, frequency, and time domain feature extraction methods.

feature extraction, time-series, statistical features, frequency domain, tsfresh, python, pandas, numpy, machine learning features

Created by Artur Dox

Perform frequency analysis on time-series signals using spectral analysis techniques. Identify dominant frequency components and analyze power spectral density.

frequency analysis, spectral analysis, FFT, time-series, python, scipy, signal processing

Created by Artur Dox

Impute missing values in time-series data using interpolation, KNN, and fill-based strategies. Evaluate imputation quality against ground truth using RMSE and MAE.

missing data, imputation, interpolation, KNN imputation, forward-fill, backward-fill, time-series, python, pandas, scikit-learn, gap filling, RMSE

Created by Artur Dox

Detect and characterize peaks and valleys in time-series signals using prominence, width, and height criteria. Aggregate peak statistics and visualize annotated results.

peak detection, valley detection, find_peaks, prominence, peak width, peak height, Savitzky-Golay smoothing, signal processing, time-series, scipy, python, amplitude distribution

Created by Artur Dox

Remove seasonal patterns from time-series using STL decomposition and X-13ARIMA-SEATS concepts. Validate adjustment quality through spectral analysis of residuals.

seasonal adjustment, STL decomposition, X-13ARIMA-SEATS, seasonality removal, loess, time-series, statsmodels, python, trend isolation, spectral analysis, calendar effects

Created by Artur Dox

Implement digital signal processing techniques for filtering and denoising time-series data. Apply various filter types and denoising algorithms.

signal filtering, denoising, digital filters, time-series, signal processing, python, scipy, Savitzky-Golay, Wiener filter

Created by Artur Dox

Detect and analyze spike synchronization patterns in neural or signal time-series data. Implement spike detection algorithms and analyze temporal relationships.

spike detection, synchronization, neural signals, time-series, python, scipy, signal processing

Created by Artur Dox

Analyze time-series data to determine stationarity and apply appropriate statistical tests and transformations. Implement stationarity tests and data transformations.

stationarity, statistical tests, time-series, transformations, python, statsmodels, ADF test

Created by Artur Dox

Compute STFT spectrograms to analyze time-varying frequency content of non-stationary signals. Explore resolution trade-offs and identify transient and harmonic structures.

STFT, Short-Time Fourier Transform, spectrogram, time-frequency analysis, windowing, Hann window, FFT, signal processing, non-stationary, time-series, scipy, python, harmonic analysis

Created by Artur Dox

Classify time-series data into categories using DTW-based nearest neighbor, ROCKET, and CNN approaches. Compare classifier performance using accuracy and F1-score.

time-series classification, DTW, Dynamic Time Warping, ROCKET, CNN, 1D convolution, sktime, tslearn, python, F1-score, confusion matrix, nearest neighbor

Created by Artur Dox

Build time-series forecasting models using statistical and machine learning approaches. Implement ARIMA, Prophet, and other forecasting techniques with performance evaluation.

forecasting, ARIMA, Prophet, time-series, prediction, python, statsmodels, fbprophet, model evaluation

Created by Artur Dox

Implement clustering algorithms specifically designed for time-series data. Group similar time-series patterns using DTW and other specialized distance metrics.

clustering, DTW, time-series, pattern recognition, similarity, python, scikit-learn, tslearn, hierarchical clustering

Created by Artur Dox

Implement time-series decomposition methods to separate trend, seasonal, and residual components. Use classical and modern decomposition techniques.

decomposition, STL, seasonal, trend, time-series, python, statsmodels, pandas, classical decomposition, X-11

Created by Artur Dox

Analyze and detect trends in time-series data using linear regression, polynomial fitting, and the Mann-Kendall statistical test with Sen's slope estimation.

trend analysis, Mann-Kendall test, linear trend, polynomial trend, Sen's slope, detrending, time-series, python, scipy, statsmodels, monotonic trend

Created by Artur Dox

Implement wavelet analysis techniques for time-frequency analysis of signals. Perform continuous and discrete wavelet transforms with scalogram visualization.

wavelet analysis, time-frequency, scalogram, signal decomposition, multi-resolution, python, pywt, scipy, CWT, DWT

Created by Artur Dox