SmartTensors AI Software for Big Data Analysis

How Unsupervised Learning Advances Big Data Analysis

Download SmartTensors, a powerful AI forecasting suite using neural networks.

SmartTensors AI software suite is an award-winning, unsupervised learning platform designed to help enterprise organizations analyze big-data, make robust predictions, and create visual presentations. Developed by Los Alamos National Laboratory, SmartTensors AI uses unsupervised learning and neural networks to find patterns, trends, and anomalies in large datasets.

The robust package also offers explainable AI, NLP, and powerful visualization tools. SmartTensors is written in Julia, Python, and C++, and operates on CPUs, GPUs, supercomputers, and personal laptops.

NASA uses SmartTensors AI to predict natural disasters.

In this video, LANL scientists explain how tensor technology, neural networks, and high-visual compression features help NASA predict the behavior of natural disasters. From asteroid-induced tsunami’s to wide-spread heat-waves, you must see to believe.

Award-Winning Features

Extract latent features in enterprise datasets.

Make robust predictions using verified data.

Detect anomalies and categorize threats.

Determine dependencies automatically.

Explain results using Natural Language Processing (NLP) and Explainable AI (XAI)

Extract and compress high-visual data: photos, videos, and satellite imagery.

Learn more about SmartTensors’ powerful tools.

Case Studies

Predicting natural disasters

NASA researchers used SmartTensors to study 3D models of asteroids. This helps them understand the hypothetical effects of tsunamis caused by asteroids. SmartTensors AI explained these realistic visual presentations in simple human language. NASA can use this information to develop future safety and contingency plans.

Classifying cybersecurity anomalies

As risk surfaces increase, it's harder to identify threats from begin anomalies. Cybersecurity teams are using SmartTensors AI to detect and classify anomalies. Even in cases of limited datasets or imbalanced data, SmartTensors can help teams verify and remediate threats.

Expanding Literature Classification

Classifying large sets of scientific literature is important for future research. However, using labeled data to build repositories does not allow a system to scale as new information is introduced. LANL used SmartTensors unsupervised learning capabilities to iteratively expand literature data sets.

NASA uses SmartTensors AI software to predict catastrophic events. LANL AI software includes image compression, XAI and NLP.

Key Capabilities & Applications

Neural Networks

Tensor Technology

Data-Filtering

Data Compression

Explainable Learning

Industry Applications

Download the SmartTensors AI Suite

Download T-ELF on GitHub

Tensor Extraction of Latent Features (T-ELF) is one of the machine learning software packages developed as part of the R&D 100 winning SmartTensors AI project at Los Alamos National Laboratory (LANL). T-ELF presents an array of customizable software solutions crafted for analysis of datasets.

Download pyCP-APR on GitHub

pyCP-APR is an award-winning Python library for breaking down and analyzing large data sets. It helps detect unusual patterns and speeds up data processing using GPUs.

Download pyDNMFk on GitHub

pyDNMFk is a software tool that breaks down large datasets using non-negative matrix factorization across multiple systems. It reduces the difference between the original and reconstructed data using different mathematical methods.

Download AdversarialTensors on GitHub

A tensor-based framework for making AI models more secure against attacks. This library uses a variety tensor methods to protect AI systems from adversarial threats.

Download pyDNTNK on GitHub

pyDNTNK is a software tool for breaking down large datasets using advanced tensor methods like Tensor Train and Hierarchical Tucker decompositions. It works across multiple systems and is built on top of pyDNMFk.

Download cuda-pyDNMFk on GitHub

Cuda-pyDNMFk is a software platform designed for breaking down large datasets and discovering hidden features. It uses Cuda and Python to handle data that is too large for regular memory processing.

Download pyDRESCALk on GitHub

pyDRESCALk is a software tool for breaking down large relational datasets using non-negative RESCAL decomposition across multiple systems.

Download pyHNMFk on GitHub

Identifying sources of advection-diffusion transport typically involves solving complex inverse models using available data. pyHNMFk simplifies this by breaking down recorded data, determining the number of unknown sources, and using the Green's function to identify their characteristics.

AI News

R&D 100 winner of the day: SmartTensors AI Platform

SmartTensors autonomously analyzes and discovers hidden features, signatures and patterns otherwise undetectable and buried in tens of terabytes of data.