Negative Learning Rates and P-Learning

نویسنده

  • Devon Merrill
چکیده

We present a method of training a differentiable function approximator for a regression task using negative examples. We effect this training using negative learning rates. We also show how this method can be used to perform direct policy learning in a reinforcement learning setting. 1 Regression and Learning Rates The goal of regression analyses is to find a regression function, a function that models the relationship between the independent variables (the inputs) and the dependent variables (the outputs). When a complex relationship between these variables, an exact regression function is not sought. Instead an approximate regression function is used to model the relationship. Function approximators have been shown to be a sound method for finding approximate regression functions. Function approximates are generally trained using example input-output pairs from the function that is to be approximated. Given the input from the pair, the output of the function approximator is compared to the actual output from the example. The function approximator is then modified –possibly through gradient descent– so that its output matches the output in the example. This is one training step. The function approximator is not usually modified so vigorously that the output matches the example output perfectly for each training. Extreme modification of the approximator tends to result in the loss or forgetting of the previous examples, that is, the previous training is overwritten by too strong of an update. Also, updates that are too vigorous tend to negatively affect the approximator’s ability to generalize to unseen examples. Another issue occures when a function approximator is updated, the parameters of the model, it are changed. If these parameters are changed too quickly, they can overrun a computer’s ability to represent these numbers. This is sometimes known as model explosion and it prevents the use of the function approximator. To solve these and other problems, the updates to a function approximator are attenuated by a fractional amount, conceptualized as the learning rate. This allows the approximator to be pushed in the desired direction by a small, tunable amount. All approximators trained by gradient descent use a learning rate. Numerous methods have been developed to automatically compute learning rates. Many use the second derivative of the parameter change (where the first derivative is the raw magnitude and direction of the update. This basic method is used from the familiar Newton’s Method, to esoteric concave optimization methods. Other methods use previous update amounts to tune the learning rate. However, the most common method for setting a learning rate is to start with 0.1 and hope for the best. 1 ar X iv :1 60 3. 08 25 3v 1 [ cs .A I] 2 7 M ar 2 01 6 2 Learning Rates as a Learning Channel In regression, the learning rate does not contain information about what the approximator should output for any given input. This information is contained in the input-output example pairs. This is why the learning rate is always positive; a positive learning rate means that the approximator should match this example more closely. Still, there are many situations where we do not have access to good input-output pairs. The pairs we have access to might not be from the actual function we are trying to model, but from any other function. The pairs we have might be totally random. However, we might have access to a measure of how closely the output we have matches up with hypothetical output, given some input, from the function we are trying to approximate. For example, take input-output pair e = (x, r) where x is some input vector and r is a random output vector. We would like our function approximator nn(x) to give us the correct output from our target function nn∗(x) = y, but we do not have access to any example nn∗(x) = y for any x. However, if we have a distance function dist∗(x, r) → R that gives a similarity measure between r and nn∗(x) = y, we can still train nn(·) We would like to use dist∗(·, ·) as a training signal to train n. Learning rate is useful as a channel for this distance training signal. To use the learning rate as a learning channel we can take each example in the training set and use it to assign a custom learning rate for each example. We should still have a global learning rate μ. Now, when we train on example i, the learning rate for example ei is μ× dist(xi, ri). Now we train a simple 2-layer artificial neural network to reproduce the sine function on the interval [−5, 5] with 40 example points. The network uses tanh activation function for the 128 hidden units. Each example is of the form ei = (xi, ri, di) where xi = −5 + i × 0.25 and ri is drawn from a uniform distribution on the interval [−1, 1]. The value di is the examples learning rate factor calculated by dist(xi) = di = |sin(xi)− ri|. Let’s see how this works in figure 1. Figure 1: Output of network trained with per example learning rates di = dist(xi, ri).

برای دانلود رایگان 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

The Effects of Social Networks on Nursing Students’ Academic Achievement and Retention in Learning English

Introduction: The use of modern virtual technologies in the process of teaching-learning is inevitable. One example is the use of virtual social networks in education. The purpose of this study was to examine the effects of social networking on nursing students’ academic achievement and retention in learning English. Methods: The pretest-posttest design with a control group was used in this qua...

متن کامل

Prediction of Iranian EFL Learners’ Learning Approaches Through Their Teachers’ Narrative Intelligence and Teaching Styles: A Structural Equation Modelling Analysis

It goes without saying that there are many influential factors affecting the success of any learning experience, and teachers are definitely among the significant factors influencing the process of teaching and learning. In this respect, the present study sought to investigate the prediction of Iranian English as a Foreign Language (EFL) learners' learning approaches through their teachers’ nar...

متن کامل

Investigating Predictors of High School Students’ Negative Attitudes Towards Learning English by Developing, Validating, and Running a Questionnaire

The purpose of this study was to explore the predictors of negative attitudes towards learning English from L2 learners’ points of view. A mixed methods research approach was adopted with a sequential exploratory design, followed by an endorsement phase. Eighteen high school students in Fars province (Iran) were interviewed on the sources of negative attitudes towards learning English. Based on...

متن کامل

Structural model of academic adjustment based on learning strategies and learning styles with the mediating role of cognitive emotion regulation strategies in university students

The purpose of the current research was to determine the mediating role of cognitive emotion regulation strategies in the relationship between learning styles and students' academic adjustment. The current research is descriptive-correlational in terms of methodology. The statistical population was all the students of Shahid Bahoner University in Kerman in the academic year of 2021-2022 (12614)...

متن کامل

The effects of traffic noise on memory and auditory-verbal learning in Persian language children

Background: Acoustic noise is one of the universal pollutants of modern society. Although the high level of noise adverse effects on human hearing has been known for many years, non-auditory effects of noise such as effects on cognition, learning, memory and reading, especially on children, have been less considered. Factors which have negative impact on these features can also have a negat...

متن کامل

Learning Curve and Industry Structure: Evidences from Iranian Manufacturing Industries

he empirical studies have shown that cost advantages can occur due to economies of scale and economies of learning. However, a few studies have attempted to distinguish between these two effects on reducing costs. This paper is the first attempt on recognizing the impact of learning on reducing the cost with distinguishing the effect of economies of scale in Iran. Therefore, this study aims to ...

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

عنوان ژورنال:
  • CoRR

دوره abs/1603.08253  شماره 

صفحات  -

تاریخ انتشار 2016