Size-Independent Sample Complexity of Neural Networks

نویسندگان

  • Noah Golowich
  • Alexander Rakhlin
  • Ohad Shamir
چکیده

We study the sample complexity of learning neural networks, by providing new bounds on their Rademacher complexity assuming norm constraints on the parameter matrix of each layer. Compared to previous work, these complexity bounds have improved dependence on the network depth, and under some additional assumptions, are fully independent of the network size (both depth and width). These results are derived using some novel techniques, which may be of independent interest.

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

Application of Artificial Neural Networks and Support Vector Machines for carbonate pores size estimation from 3D seismic data

This paper proposes a method for the prediction of pore size values in hydrocarbon reservoirs using 3D seismic data. To this end, an actual carbonate oil field in the south-western part ofIranwas selected. Taking real geological conditions into account, different models of reservoir were constructed for a range of viable pore size values.  Seismic surveying was performed next on these models. F...

متن کامل

A Laboratory Study on Stress Dependency of Joint Transmissivity and its Modeling with Neural Networks, Fuzzy Method and Regression Analysis

Correct estimation of water inflow into underground excavations can decrease safety risks and associated costs. Researchers have proposed different methods to asses this value. It has been proved that water transmissivity of a rock joint is a function of factors, such as normal stress, joint roughness and its size and water pressure therefore, a laboratory setup was proposed to quantitatively m...

متن کامل

Computational Learning Theory for Artificial Neural Networks

There are many types of activity which are commonly known as ‘learning’. Here, we shall discuss a mathematical model of one such process, known as the the ‘probably approximately correct’ (or PAC) model. We shall illustrate how key problems of learning in artificial neural networks can be studied within this framework, presenting theoretical analyses of two important issues: the size of trainin...

متن کامل

Diagnosis of brain tumor using image processing and determination of its type with RVM neural networks

Typically, the diagnosis of a tumor is done through surgical sampling, which is more precise with existing methods. The difference is that this is an aggressive, time consuming and expensive way. In the statistical method, due to the complexity of the brain tissues and the similarity between the cancerous cells and the natural tissues, even a radiologist or an expert physician may also be in er...

متن کامل

Rapid and Simultaneous Determination of Montelukast, Fexofenadine and Cetirizine Using Partial Least Squares and Artificial Neural Networks Modeling

Simultaneous determination of pharmaceutical compounds and accurate quantitative prediction of them are of great interest in the clinical and laboratory-based investigations.This work has focused on a comprehensive comparison of Partial Least-Squares (PLS-1) and Artificial Neural Networks (ANN) as two powerful types of chemometric methods. For this purpose, montelukast (MONT), fexofenadine ...

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:
  • CoRR

دوره abs/1712.06541  شماره 

صفحات  -

تاریخ انتشار 2017