Large-scale probabilistic prediction with and without validity guarantees

نویسندگان

  • Vladimir Vovk
  • Ivan Petej
  • Valentina Fedorova
چکیده

This paper studies theoretically and empirically a method of turning machinelearning algorithms into probabilistic predictors that automatically enjoys a property of validity (perfect calibration) and is computationally e cient. The price to pay for perfect calibration is that these probabilistic predictors produce imprecise (in practice, almost precise for large data sets) probabilities. When these imprecise probabilities are merged into precise probabilities, the resulting predictors, while losing the theoretical property of perfect calibration, are consistently more accurate than the existing methods in empirical studies.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Large-scale probabilistic predictors with and without guarantees of validity

This paper studies theoretically and empirically a method of turning machinelearning algorithms into probabilistic predictors that automatically enjoys a property of validity (perfect calibration) and is computationally efficient. The price to pay for perfect calibration is that these probabilistic predictors produce imprecise (in practice, almost precise for large data sets) probabilities. Whe...

متن کامل

Probabilistic View of Occurrence of Large Earthquakes in Iran

In this research seismicity parameters, repeat times and occurrence probability of large earthquakes are estimated for 35 seismic lineaments in Persian plateau and the surrounding area. 628 earthquakes of historical time and present century with MW>5.5 were used for further data analysis. A probabilistic model is used for forecasting future large earthquake occurrences in each chosen lineament....

متن کامل

Prediction of Risk of Falling among Institutionalized Elderly People in Iran

Aging declines abilities and leads to increased risk of falling and subsequently poorer quality of life. The objective of this study was to identify a proper assessment for risk of falling among institutionalized elderly. In order to accomplish this aim, two functional assessment methods, Berg Scale and Tinetti Scale, were used and the validity indices of these methods were calculated. Prospect...

متن کامل

LPKP: location-based probabilistic key pre-distribution scheme for large-scale wireless sensor networks using graph coloring

Communication security of wireless sensor networks is achieved using cryptographic keys assigned to the nodes. Due to resource constraints in such networks, random key pre-distribution schemes are of high interest. Although in most of these schemes no location information is considered, there are scenarios that location information can be obtained by nodes after their deployment. In this paper,...

متن کامل

Application of Probabilistic Clustering Algorithms to Determine Mineralization Areas in Regional-Scale Exploration Studies

In this work, we aim to identify the mineralization areas for the next exploration phases. Thus, the probabilistic clustering algorithms due to the use of appropriate measures, the possibility of working with datasets with missing values, and the lack of trapping in local optimal are used to determine the multi-element geochemical anomalies. Four probabilistic clustering algorithms, namely PHC,...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2015