NLP Time Series Encoding for ML Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine-learning models often fail to accurately analyze time series data due to their inability to capture contextual information, leading to less accurate predictions and increased processing power consumption.
Innovation Solution
Applying natural language processing (NLP) techniques to time series data by generating encodings that are treated similarly to text, allowing NLP models to derive contextual attributes which are then used to improve the accuracy and efficiency of machine-learning models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional machine-learning models are used to analyze time series data, then the models can process the data, but they fail to capture contextual information leading to less accurate predictions
Solution Approach 1:
The patent introduces NLP models as an intermediary between traditional machine-learning models and time series data. The NLP models process time series data to extract contextual attributes (such as trend, seasonality, and anomaly detection) which are then fed into the machine-learning model, enabling it to capture contextual information that would otherwise be lost
Solution Approach 2:
The patent transforms time series data into different parameter representations by applying NLP techniques. The data is converted into various formats including tokenized sequences, n-grams, and contextual embeddings, which preserve contextual information while maintaining compatibility with machine-learning models
2Productivity
If traditional machine-learning models process time series data without NLP techniques, then the processing is simpler, but the models consume more processing power and time
Solution Approach 1:
The patent segments the data processing task into two distinct stages: first, NLP models extract contextual attributes and features from raw time series data; second, these extracted features are fed into the machine-learning model for final prediction. This segmentation allows each model to operate on optimized data, reducing overall processing power consumption
Solution Approach 2:
The patent applies preliminary processing using NLP techniques to extract and preprocess contextual attributes before the main machine-learning model processes the data. This preliminary action of extracting trends, seasonality, and anomalies upfront reduces the computational burden on the subsequent prediction stage
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Techniques are described herein for applying natural language processing (NLP) techniques to time series data to derive attributes of an object for use with a machine-learning model. In one example, a system can receive a time series associated with an object over a time window, where the time series includes a set of discrete values. The system can then generate a time series encoding based on the time series. The system can provide the time series encoding as input to a trained natural language processing (NLP) model, which can generate one or more output embeddings based on the time series encoding. Next, the system can determine at least one attribute associated with the object based on the one or more output embeddings. The system can then provide the attributes for use with a machine-learning model, which may for example be configured to predict a future characteristic of the object.