Neural network for tabular data
AssigneeApplied Materials Israel Ltd.
InventorHyung Geun LEE
Life of this patent
Estimating the term…
Abstract
The presently disclosed subject matter includes a novel computer-implemented method and computer system for the classification of tabular data using a new neural network classifier model (also referred to herein as “Tabular Neural Network Classifier” or TNNC). The disclosed method and system are characterized by improved accuracy and efficiency, as compared to other existing tabular data classification techniques such as Random Forests, XGBoost, etc. The inventor found that the TNNC exhibits in general a better TP to FP ratio in the classification output and a shorter processing time, as compared to existing tabular data classification techniques.
Claims20
A computer system comprising at least one processing circuitry configured to classify tabular data using a machine learning model, the processing circuitry being configured to:
- obtain tabular data comprising one or more records, each record corresponding to a respective sample and comprising multiple features characterizing the sample;
- utilize a machine learning (ML) model on the tabular data, wherein the ML model comprises a collective input layer, a collective integration layer, and multiple neural network (NN) threads, each NN thread connected between the collective input layer and the collective integration layer and being designed as an individual neural network (NN) comprising at least two hidden layers and an output layer;
- wherein the processing circuitry is configured for utilizing the ML model to:
- provide to the ML model a collection of input features extracted from the tabular data;
- for each NN thread:
- apply a dropout operation on the collection of input features, to thereby obtain a respective subset of features selected from the collection input features;
- apply a neural network algorithm on the respective subset of features to thereby obtain respective NN thread output at a respective output layer;
- and
- provide the respective NN thread output to the collective integration layer;
- calculate, at the collective integration layer, a collective output of the ML model based on the multiple NN outputs, the collective output being indicative of a classification of the sample to a class selected from at least two classes.
A computer implemented method of classification of tabular data using a deep learning ML model, the method comprising:
- obtaining tabular data comprising one or more records, each record corresponding to a respective sample and comprising multiple features characterizing the sample;
- utilizing a machine learning (ML) model on the tabular data, wherein the ML model comprises a collective input layer, a collective integration layer, and multiple neural network (NN) threads, each NN thread connected between the collective input layer and the collective integration layer and being designed as an individual neural network (NN) comprising at least two hidden layers and an output layer;
- wherein utilizing the ML model comprises:
- providing to the ML model a collection of input features extracted from the tabular data;
- for each NN thread:
- applying a dropout operation on the collection of input features, to thereby obtain a respective subset of features selected from the collection input features;
- applying a neural network algorithm on the respective subset of features to thereby obtain respective NN thread output at a respective output layer;
- and
- providing the respective NN thread output to the collective integration layer;
- calculating, at the collective integration layer, a collective output of the ML model based on the multiple NN outputs, the collective output being indicative of a classification of the sample to a class selected from at least two classes.
A non-transitory computer readable medium comprising instructions that, when executed by a computer, cause the computer to perform a computerized method of classification of tabular data using a machine learning (ML) model, the method comprising:
- obtaining tabular data comprising one or more records, each record corresponding to a respective sample and comprising multiple features characterizing the sample;
- utilizing a machine learning (ML) model on the tabular data, wherein the ML model comprises a collective input layer, a collective integration layer, and multiple neural network (NN) threads, each NN thread connected between the collective input layer and the collective integration layer and being designed as an individual neural network (NN) comprising at least two hidden layers and an output layer;
- wherein utilizing the ML model comprises:
- providing to the ML model a collection of input features extracted from the tabular data;
- for each NN thread:
- applying a dropout operation on the collection of input features, to thereby obtain a respective subset of features selected from the collection input features;
- applying a neural network algorithm on the respective subset of features to thereby obtain respective NN thread output at a respective output layer;
- and
- providing the respective NN thread output to the collective integration layer;
- calculating, at the collective integration layer, a collective output of the ML model based on the multiple NN outputs, the collective output being indicative of a classification of the sample to a class selected from at least two classes.
Dependencies are inferred from the claim text.