Using Stacking Approaches for Machine Learning Models

@article{Pavlyshenko2018UsingSA,
  title={Using Stacking Approaches for Machine Learning Models},
  author={Bohdan M. Pavlyshenko},
  journal={2018 IEEE Second International Conference on Data Stream Mining \& Processing (DSMP)},
  year={2018},
  pages={255-258}
}
  • B. Pavlyshenko
  • Published 1 August 2018
  • Computer Science
  • 2018 IEEE Second International Conference on Data Stream Mining & Processing (DSMP)
In this paper, we study the usage of stacking approach for building ensembles of machine learning models. The cases for time series forecasting and logistic regression have been considered. The results show that using stacking technics we can improve performance of predictive models in considered cases. 

Figures from this paper

Machine-Learning Models for Sales Time Series Forecasting
TLDR
The results show that using stacking techniques, a stacking approach for building regression ensemble of single models, can improve the performance of predictive models for sales time series forecasting.
Time Series Forecasting Using Machine Learning
TLDR
A time series forecasting approach is used with machine learning techniques to forecast the store item demands to improve accuracy, throughput and overall performance.
Short-Term Prediction of Global Solar Radiation Energy Using Weather Data and Machine Learning Ensembles: A Comparative Study
The ability to predict solar radiation one-day-ahead is critical for the best management of renewable energy tied-grids. Several machine learning ensemble techniques have been proposed to enhance
Stacking-Based Ensemble of Support Vector Regressors for One-Day Ahead Solar Irradiance Prediction
TLDR
Several ensemble methods based on machine learning named stacking for one-day-ahead predicting of solar radiation intensity are surveyed and the stacking based ensemble models outperformed all individual models as well as other combining techniques.
Can ensemble machine learning be used to predict the groundwater level dynamics of farmland under future climate: a 10-year study on Huaibei Plain
Accurate and simple prediction of farmland groundwater level (GWL) is an important aspect of agricultural water management. A farmland GWL prediction model, GWPRE, was developed that integrates four
Sales Time Series Analytics Using Deep Q-Learning
TLDR
Results show that using deep Q-learning, the decision making process for price optimization and supply-demand problems can be optimized and the trained agent can be used in real business environment.
A Model Stacking Approach for Ride-Hailing Demand Forecasting : a Case Study of Algiers
TLDR
This work harnesses the power of statistical and machine learning models in a joint ensemble model and proposes a stacking approach which combines the predictions of ARIMA, SARIMa, MLP, LSTM and XGBoost, and considers also external factors such weather conditions and national holidays.
Automated Evolutionary Approach for the Design of Composite Machine Learning Pipelines
A Calibrated Combination of Probabilistic Precipitation Forecasts to Achieve a Seamless Transition from Nowcasting to Very Short-Range Forecasting
TLDR
A new model for the combination of two or more probabilistic forecasts is presented, based on a logit transformation of the underlying initiator of the forecasts, which shows promising results in terms of predictability and uncertainty.
Diversity as The Basis for Effective Clustering-Based Classification
TLDR
An approach based on clustering is proposed, according to which the influence of an individual model is inversely proportional to the volumes of aggregated groups, which improves the possibility of building an ensemble based on the properties of the correlation parameters of the models.
...
1
2
3
4
...

References

SHOWING 1-7 OF 7 REFERENCES
Linear, machine learning and probabilistic approaches for time series analysis
  • B. Pavlyshenko
  • Computer Science
    2016 IEEE First International Conference on Data Stream Mining & Processing (DSMP)
  • 2016
TLDR
The forecasting approaches using linear models, ARIMA algorithm, XGBoost machine learning algorithm, and probabilistic modeling the approaches using copulas and Bayesian inference are considered.
Machine learning, linear and Bayesian models for logistic regression in failure detection problems
  • B. Pavlyshenko
  • Computer Science
    2016 IEEE International Conference on Big Data (Big Data)
  • 2016
TLDR
This work used the data from Kaggle competition “Bosch Production Line Performance” as a data set for the analysis and considered the use of machine learning, linear and Bayesian models for logistic regression in manufacturing failures detection.
XGBoost: A Scalable Tree Boosting System
TLDR
This paper proposes a novel sparsity-aware algorithm for sparse data and weighted quantile sketch for approximate tree learning and provides insights on cache access patterns, data compression and sharding to build a scalable tree boosting system called XGBoost.
Stacked generalization
Greedy function approximation: A gradient boosting machine.
TLDR
A general gradient descent boosting paradigm is developed for additive expansions based on any fitting criterion, and specific algorithms are presented for least-squares, least absolute deviation, and Huber-M loss functions for regression, and multiclass logistic likelihood for classification.
Stochastic gradient boosting
Doing Bayesian Data Analysis: A Tutorial with R, JAGS, and Stan
TLDR
Doing Bayesian Data Analysis: A Tutorial with R, JAGS, and Stan provides an accessible approach to Bayesian data analysis, as material is explained clearly with concrete examples.