18062673. GENERATING CAUSAL ASSOCIATION RANKINGS USING DYNAMIC EMBEDDINGS simplified abstract (International Business Machines Corporation)
Contents
GENERATING CAUSAL ASSOCIATION RANKINGS USING DYNAMIC EMBEDDINGS
Organization Name
International Business Machines Corporation
Inventor(s)
Jiri Navratil of Cortlandt Manor NY (US)
Karthikeyan Shanmugam of Bengaluru (IN)
Youssef Mroueh of New York NY (US)
Inkit Padhi of White Plains NY (US)
GENERATING CAUSAL ASSOCIATION RANKINGS USING DYNAMIC EMBEDDINGS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18062673 titled 'GENERATING CAUSAL ASSOCIATION RANKINGS USING DYNAMIC EMBEDDINGS
The embodiment described in the abstract is a system for generating causal association rankings for candidate events within a window of candidate events using dynamic deep neural network generated embeddings.
- Automatically receives a window of candidate events, including events of a first type preceding target events of interest.
- Generates contrastive windows of candidate events, each corresponding to a different dropped candidate event from the received window.
- Identifies matching historical windows of events with embeddings close in distance to the embeddings of the contrastive windows and calculates a first score for each match.
- Identifies matching incident windows and calculates a corresponding second score.
- Uses the first and second scores to generate causal association rankings.
- Potential Applications:**
- Predictive analytics in various industries such as finance, healthcare, and marketing. - Anomaly detection in cybersecurity. - Recommendation systems for e-commerce platforms.
- Problems Solved:**
- Efficiently identifying causal relationships between events. - Improving the accuracy of predictive models. - Enhancing decision-making processes based on historical data.
- Benefits:**
- Increased accuracy in predicting future events. - Better understanding of the relationships between different events. - Enhanced performance of machine learning algorithms.
- Commercial Applications:**
- Predictive maintenance in manufacturing. - Fraud detection in financial services. - Personalized recommendations in online retail.
- Prior Art:**
Prior research in the field of causal inference and event prediction using neural networks.
- Frequently Updated Research:**
Ongoing studies on improving the efficiency and accuracy of causal association rankings in dynamic environments.
- Questions about the Technology:**
1. How does the system determine the significance of the causal relationships between events? 2. What are the potential limitations of using dynamic deep neural network generated embeddings for event prediction and causal inference?
Original Abstract Submitted
An embodiment for generating causal association rankings for candidate events within a window of candidate events using dynamic deep neural network generated embeddings. The embodiment may automatically receive a window of candidate events including events of a first type preceding one or more target events of interest. The embodiment may automatically generate contrastive windows of candidate events, each of the contrastive windows of candidate events of the first type corresponding to a different dropped candidate event from the received window of candidate events. The embodiment may automatically identify matching historical windows of events having resulting embeddings that are close in distance to the embeddings corresponding to the embeddings of the contrastive windows and calculate a first score for each match. The embodiment may automatically identify matching incident windows and calculate a corresponding second score. The embodiment may use the first and second scores to generate casual association rankings.
