3. Data, Science & AI

Monitors Teradata query execution in real-time, analyzes performance patterns, and provides insights from query logs.

Performs missing value imputation on tabular datasets using the TD_SimpleImputeFit algorithm.

Provides windowing functions to reduce spectral leakage in signal analysis, enhancing accuracy in Fourier transform-based spectral processing.

Performs statistical outlier detection and handling on datasets using the TD_OutlierFit algorithm for data quality assurance.

Provides a Naive Bayes classifier for probabilistic classification tasks in machine learning applications.

Provides a decision tree classifier for categorical prediction tasks and extracts interpretable rules from the model.

Analyzes sequential event data to identify recurring patterns and trends in user or system behavior.

Performs ARIMA time series forecasting for predictive analytics and trend analysis.

Performs time series differencing to achieve stationarity and remove trends, essential for time series modeling and analysis.

Performs linear regression analysis to predict continuous numerical outcomes from input features.

Provides digital filtering capabilities for noise reduction and signal enhancement in data processing tasks.

Generates smoothed time series predictions using moving average algorithms on historical data.

Performs signal resampling and interpolation to convert signal rates, essential for audio and sensor data processing.

Performs batch OCR on images using DeepSeek-OCR via Ollama, converting visual text to machine-readable data.

Executes Ljung-Box portmanteau tests to diagnose time series models by analyzing residual autocorrelation.

Provides time series-specific cross-validation techniques for validating machine learning models on sequential data.

Performs K-means clustering to segment customers and group data points based on similarity.

Provides a high-accuracy random forest ensemble classifier for machine learning classification tasks.

Provides categorical variable encoding via TD_OneHotEncodingFit and Transform methods for machine learning data preparation.

Monitors active Teradata sessions in real-time, providing SQL execution details, blocking issue identification, and optional control actions.

Estimates parameters for ARIMA, ARMA, and seasonal time series models, supporting both non-seasonal and seasonal variants.

Analyzes time-series data to extract frequency domain features and detect periodic patterns using power spectrum techniques.

Performs Fourier transformation for frequency domain analysis, enabling spectral decomposition of signals and data.

Identifies structural breaks in time series data by detecting significant shifts in statistical properties for anomaly detection and trend analysis.