اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدDebiasing Stochastic Gradient Descent to handle missing values
407
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Stochastic gradient algorithm is a key ingredient of many machine learning methods, particularly appropriate for large-scale learning.However, a major caveat of large data is their incompleteness.We propose an averaged stochastic gradient algorithm handling missing values in linear models. This approach has the merit to be free from the need of any data distribution modeling and to account for heterogeneous missing proportion.In both streaming and finite-sample settings, we prove that this algorithm achieves convergence rate of $mathcal{O}(frac{1}{n})$ at the iteration $n$, the same as without missing values. We show the convergence behavior and the relevance of the algorithm not only on synthetic data but also on real data sets, including those collected from medical register.
قيم البحث
اقرأ أيضاً
In many biomedical applications, outcome is measured as a ``time-to-event (eg. disease progression or death). To assess the connection between features of a patient and this outcome, it is common to assume a proportional hazards model, and fit a prop
ortional hazards regression (or Cox regression). To fit this model, a log-concave objective function known as the ``partial likelihood is maximized. For moderate-sized datasets, an efficient Newton-Raphson algorithm that leverages the structure of the objective can be employed. However, in large datasets this approach has two issues: 1) The computational tricks that leverage structure can also lead to computational instability; 2) The objective does not naturally decouple: Thus, if the dataset does not fit in memory, the model can be very computationally expensive to fit. This additionally means that the objective is not directly amenable to stochastic gradient-based optimization methods. To overcome these issues, we propose a simple, new framing of proportional hazards regression: This results in an objective function that is amenable to stochastic gradient descent. We show that this simple modification allows us to efficiently fit survival models with very large datasets. This also facilitates training complex, eg. neural-network-based, models with survival data.
Nonparametric regression with random design is considered. Estimates are defined by minimzing a penalized empirical $L_2$ risk over a suitably chosen class of neural networks with one hidden layer via gradient descent. Here, the gradient descent proc
edure is repeated several times with randomly chosen starting values for the weights, and from the list of constructed estimates the one with the minimal empirical $L_2$ risk is chosen. Under the assumption that the number of randomly chosen starting values and the number of steps for gradient descent are sufficiently large it is shown that the resulting estimate achieves (up to a logarithmic factor) the optimal rate of convergence in a projection pursuit model. The final sample size performance of the estimates is illustrated by using simulated data.
Despite the strong theoretical guarantees that variance-reduced finite-sum optimization algorithms enjoy, their applicability remains limited to cases where the memory overhead they introduce (SAG/SAGA), or the periodic full gradient computation they
require (SVRG/SARAH) are manageable. A promising approach to achieving variance reduction while avoiding these drawbacks is the use of importance sampling instead of control variates. While many such methods have been proposed in the literature, directly proving that they improve the convergence of the resulting optimization algorithm has remained elusive. In this work, we propose an importance-sampling-based algorithm we call SRG (stochastic reweighted gradient). We analyze the convergence of SRG in the strongly-convex case and show that, while it does not recover the linear rate of control variates methods, it provably outperforms SGD. We pay particular attention to the time and memory overhead of our proposed method, and design a specialized red-black tree allowing its efficient implementation. Finally, we present empirical results to support our findings.
The goal of regression is to recover an unknown underlying function that best links a set of predictors to an outcome from noisy observations. In non-parametric regression, one assumes that the regression function belongs to a pre-specified infinite
dimensional function space (the hypothesis space). In the online setting, when the observations come in a stream, it is computationally-preferable to iteratively update an estimate rather than refitting an entire model repeatedly. Inspired by nonparametric sieve estimation and stochastic approximation methods, we propose a sieve stochastic gradient descent estimator (Sieve-SGD) when the hypothesis space is a Sobolev ellipsoid. We show that Sieve-SGD has rate-optimal MSE under a set of simple and direct conditions. We also show that the Sieve-SGD estimator can be constructed with low time expense, and requires almost minimal memory usage among all statistically rate-optimal estimators, under some conditions on the distribution of the predictors.
We propose a new algorithm called Parle for parallel training of deep networks that converges 2-4x faster than a data-parallel implementation of SGD, while achieving significantly improved error rates that are nearly state-of-the-art on several bench
marks including CIFAR-10 and CIFAR-100, without introducing any additional hyper-parameters. We exploit the phenomenon of flat minima that has been shown to lead to improved generalization error for deep networks. Parle requires very infrequent communication with the parameter server and instead performs more computation on each client, which makes it well-suited to both single-machine, multi-GPU settings and distributed implementations.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
