Automatic Data Collection
Automatic collection and enrichment of massive of financial and non-financial data of companies
The software, due to the automatic generation of data arrays from various open sources, allows improving machine learning models. Moreover, arrays are analyzed for abnormal values and outliers
System requirements include the ability to: divide the formed massive into homogeneous groups based on the AverageKNN filtering method, form additional synthetic data to restore the balance of covariates using the SMOTE (Synthetic Minority Oversampling Technique) method, detect abnormal data samples using the method proposed by Tomek
- Boosting your machine models
- Generation of data sets
- Validation of data set values
- Formation of synthetic data sets
Key product highlights
Adaptation of data flows to fetch the segment you need most combined with possibility of data enrichment to deal with the missing values
- Automated ML-powered search of data streams
- Reliable storage of your data
- Tailoring fetching algorithm per client need
- Adding extra data from own data sets
