Data preparation is the first and a very important step towards any Large Language Model (LLM) development. This paper introduces an easy-to-use, extensible, and scale-flexible open-source data preparation toolkit called Data Prep Kit (DPK). DPK is architected and designed to enable users to scale their data preparation to their needs. With DPK they can prepare data on a local machine or effortlessly scale to run on a cluster with thousands of CPU Cores. DPK comes with a highly scalable, yet extensible set of modules that transform natural language and code data. If the user needs additional transforms, they can be easily developed using extensive DPK support for transform creation. These modules can be used independently or pipelined to perform a series of operations. In this paper, we describe DPK architecture and show its performance from a small scale to a very large number of CPUs. The modules from DPK have been used for the preparation of Granite Models [1] [2]. We believe DPK is a valuable contribution to the AI community to easily prepare data to enhance the performance of their LLM models or to fine-tune models with Retrieval-Augmented Generation (RAG).

A large number of time series forecasting models including traditional statistical models, machine learning models and more recently deep learning have been proposed in the literature. However, choosing the right model along with good parameter values that performs well on a given data is still challenging. Automatically providing a good set of models to users for a given dataset saves both time and effort from using trial-and-error approaches with a wide variety of available models along with parameter optimization. We present AutoAI for Time Series Forecasting (AutoAI-TS) that provides users with a zero configuration (zero-conf) system to efficiently train, optimize and choose best forecasting model among various classes of models for the given dataset. With its flexible zero-conf design, AutoAI-TS automatically performs all the data preparation, model creation, parameter optimization, training and model selection for users and provides a trained model that is ready to use. For given data, AutoAI-TS utilizes a wide variety of models including classical statistical models, Machine Learning (ML) models, statistical-ML hybrid models and deep learning models along with various transformations to create forecasting pipelines. It then evaluates and ranks pipelines using the proposed T-Daub mechanism to choose the best pipeline. The paper describe in detail all the technical aspects of AutoAI-TS along with extensive benchmarking on a variety of real world data sets for various use-cases. Benchmark results show that AutoAI-TS, with no manual configuration from the user, automatically trains and selects pipelines that on average outperform existing state-of-the-art time series forecasting toolkits.

Time series value forecasting using machine learning models utilizing time series features has recently got good attention of Time series analytics community. This paper proposes an automated feature learning mechanisms to filter out most useful features from hundreds of available features for time series prediction problems. The paper further proposes a novel mechanism to dynamically filter features that are most suitable for the given input time series data. With such mechanisms we create pipeline consisting of most useful features for given input data and increases the performance of the prediction model. Our proposed mechanism first, groups well known features for time series analysis, generates and assigns the features importance score using multiple scoring configurations. Once scores are assigned, features are filtered using a threshold that is derived using reference feature score and Critical Difference diagram. The filtered features are subsequently analyzed based on the characteristics of the input dataset. We show using experimental results that our approach of input data based dynamic feature selection improves the overall performance of machine learning models compared to the case where dynamic feature extraction is not applied prior to modeling.

Discovering temporal lagged and inter-dependencies in multivariate time series data is an important task. However, in many real-world applications with big data, such as commercial cloud management or predictive maintenance in manufacturing, such dependencies can be time-variant and non-linear, which makes it more challenging to extract such dependencies through traditional methods like Granger causality or statistical models. In this work, we present a novel deep learning model that uses multiple layers of adapted gated recurrent units (GRUs) for discovering both time lagged behaviors and inter-timeseries dependencies, representing them in the form of directed weighted graphs. Each individual time series is first analyzed by a pair of encoding-decoding GRUs in order to discover the time lagged dependencies and representing its samples as high dimensional vectors. Such vectors collected from all component time series are then analyzed by a decoding network component to discover inter-dependencies across all time series while forecasting their next values in the multivariate time series. Though the discovery of two types of dependencies are separated at two levels of our neural network, they are tightly connected and jointly trained in an end-to-end manner. With this joint training, improvement in learning of one type of dependency immediately impacts the learning process of the other one, leading to the overall highly accurate dependencies discovery. We empirically test our model on synthetic time series data in which the exact form of dependencies are known. We also practically evaluate its performance on two real-world applications, (i) dynamic multivariate performance monitoring data with high volatility from a commercial cloud provider and, (ii) multivariate time series generated by sensors for a manufacturing plant. We show how our approach is capable of capturing these dependency behaviors via intuitive and interpretable dependency graphs and use them to generate forecasting values.

Change detection in system behavior and its root cause detection is essential for many large-scale systems such as, manufacturing plants, in order to keep systems running uninterrupted and avoid costly machine breakdown via predictive maintenance. In this paper, we present a novel graph based technique that uses time variant interdependencies and lagged dependencies among different components of a system to detect changes in the system behavior. We further find the root causes for these detected changes by pointing out the component and its historical values that are responsible for initiating and changing the system to the new state. The proposed mechanism extracts these time variant dependencies using a deep learning system and converts them into weighted directed graphs and applies graph based techniques for change detection. For each detected change, our system uses graph theoretic techniques to uncover the root causes for the change. Such a mechanism provides us with valuable insights about the inner workings of a system from a different perspective as opposed to traditional techniques for root cause analysis that directly apply statistical models to the time series data for analysis. Experimental results on real manufacturing data show, that we can detect changes in system behavior and accurately identify the root causes in almost 71% of the cases for which we have the ground truth. For synthetic data, our system can correctly identify root causes in 87% of the cases.

Performance monitoring of cloud-native applications that consist of several micro-services involves the analysis of time series data collected from the infrastructure, platform, and application layers of the cloud software stack. The analysis of the runtime dependencies amongst the component microservices is an essential step towards performing cloud resource management, detecting anomalous behavior of cloud applications, and meeting customer Service Level Agreements (SLAs). Finding such dependencies is challenging due to the non-linear nature of interactions, aberrant data measurements and lack of domain knowledge. In this paper, we propose a novel use of the modeling capability of Long-Short Term Memory (LSTM) recurrent neural networks, which excel in capturing temporal relationships in multi-variate time series data and being resilient to noisy pattern representations. Our proposed technique looks into the LSTM model structure, to uncover dependencies amongst performance metrics, which were learned during training. We further apply this technique in three monitoring use cases, namely finding the strongest performance predictors, discovering lagged/temporal dependencies, and improving the accuracy of forecasting for a given metric. We demonstrate the viability of our approach, by comparing the results of our proposed method in the three use cases with those obtained from previously proposed methods, such as Granger causality and the classical statistical time series analysis models, such as ARIMA and Holt-Winters. For our experiments and analysis, we use performance monitoring data collected from two sources: a controlled experiment involving a sample cloud application that we deployed in a public cloud infrastructure and cloud monitoring data collected from the monitoring service of an operational, public cloud service provider.

Apache Spark enables fast computations and greatly accelerates analytics applications by efficiently utilizing the main memory and caching data for later use. At its core Apache Spark uses data structures called RDDs (Resilient Distributed Datasets) to give a unified view to the distributed data. However, the data represented in the RDDs remain unencrypted which can result in leakage of confidential data produced or processed by applications. Apache Spark persists (unencrypted) RDDs to the disk storage under various circumstances including but not limited to caching, RDD checkpointing and data spill during the data shuffling operations, etc. This lack of security makes Apache Spark unsuitable for processing of sensitive information that should be secured at all times. Moreover, RDDs stored in the main memory are prone to main-memory attacks such as RAM-scrapping. In this paper, we propose and develop solutions to fill-up such security lapses in the current Apache Spark framework. We present three different approaches to incorporate security in the Apache Spark framework. These approaches are designed to limit the exposure of unencrypted data during data processing, caching and data spill to disk. We use combination of cryptographic splitting and encryption to secure data stored and spilled by Apache Spark, both to the disk as well as to the main memory. Our approaches provide strong security by incorporating combination of Information Dispersal Algorithm (IDA) and Shamir's Perfect Secret Sharing (PSS). Extensive experimentation show that with appropriately chosen parameters our security approaches provide high security at a performance penalty between 10%-25%.