اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدHow Much Automation Does a Data Scientist Want?
122
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Data science and machine learning (DS/ML) are at the heart of the recent advancements of many Artificial Intelligence (AI) applications. There is an active research thread in AI, autoai, that aims to develop systems for automating end-to-end the DS/ML Lifecycle. However, do DS and ML workers really want to automate their DS/ML workflow? To answer this question, we first synthesize a human-centered AutoML framework with 6 User Role/Personas, 10 Stages and 43 Sub-Tasks, 5 Levels of Automation, and 5 Types of Explanation, through reviewing research literature and marketing reports. Secondly, we use the framework to guide the design of an online survey study with 217 DS/ML workers who had varying degrees of experience, and different user roles matching to our 6 roles/personas. We found that different user personas participated in distinct stages of the lifecycle -- but not all stages. Their desired levels of automation and types of explanation for AutoML also varied significantly depending on the DS/ML stage and the user persona. Based on the survey results, we argue there is no rationale from user needs for complete automation of the end-to-end DS/ML lifecycle. We propose new next steps for user-controlled DS/ML automation.
قيم البحث
اقرأ أيضاً
In order to stabilize the behavior of noisy systems, confining it around a desirable state, an effort is required to suppress the intrinsic noise. This noise suppression task entails a cost. For the important case of thermal noise in an overdamped sy
stem, we show that the minimum cost is achieved when the system control parameters are held constant: any additional deterministic or random modulation produces an increase of the cost. We discuss the implications of this phenomenon for those overdamped systems whose control parameters are intrinsically noisy, presenting a case study based on the example of a Brownian particle optically trapped in an oscillating potential.
Locally checkable labeling problems (LCLs) are distributed graph problems in which a solution is globally feasible if it is locally feasible in all constant-radius neighborhoods. Vertex colorings, maximal independent sets, and maximal matchings are e
xamples of LCLs. On the one hand, it is known that some LCLs benefit exponentially from randomness---for example, any deterministic distributed algorithm that finds a sinkless orientation requires $Theta(log n)$ rounds in the LOCAL model, while the randomized complexity of the problem is $Theta(log log n)$ rounds. On the other hand, there are also many LCLs in which randomness is useless. Previously, it was not known if there are any LCLs that benefit from randomness, but only subexponentially. We show that such problems exist: for example, there is an LCL with deterministic complexity $Theta(log^2 n)$ rounds and randomized complexity $Theta(log n log log n)$ rounds.
It is observed in the literature that data augmentation can significantly mitigate membership inference (MI) attack. However, in this work, we challenge this observation by proposing new MI attacks to utilize the information of augmented data. MI att
ack is widely used to measure the models information leakage of the training set. We establish the optimal membership inference when the model is trained with augmented data, which inspires us to formulate the MI attack as a set classification problem, i.e., classifying a set of augmented instances instead of a single data point, and design input permutation invariant features. Empirically, we demonstrate that the proposed approach universally outperforms original methods when the model is trained with data augmentation. Even further, we show that the proposed approach can achieve higher MI attack success rates on models trained with some data augmentation than the existing methods on models trained without data augmentation. Notably, we achieve a 70.1% MI attack success rate on CIFAR10 against a wide residual network while the previous best approach only attains 61.9%. This suggests the privacy risk of models trained with data augmentation could be largely underestimated.
Algorithms often have tunable parameters that impact performance metrics such as runtime and solution quality. For many algorithms used in practice, no parameter settings admit meaningful worst-case bounds, so the parameters are made available for th
e user to tune. Alternatively, parameters may be tuned implicitly within the proof of a worst-case approximation ratio or runtime bound. Worst-case instances, however, may be rare or nonexistent in practice. A growing body of research has demonstrated that data-driven algorithm design can lead to significant improvements in performance. This approach uses a training set of problem instances sampled from an unknown, application-specific distribution and returns a parameter setting with strong average performance on the training set. We provide a broadly applicable theory for deriving generalization guarantees that bound the difference between the algorithms average performance over the training set and its expected performance. Our results apply no matter how the parameters are tuned, be it via an automated or manual approach. The challenge is that for many types of algorithms, performance is a volatile function of the parameters: slightly perturbing the parameters can cause large changes in behavior. Prior research has proved generalization bounds by employing case-by-case analyses of greedy algorithms, clustering algorithms, integer programming algorithms, and selling mechanisms. We uncover a unifying structure which we use to prove extremely general guarantees, yet we recover the bounds from prior research. Our guarantees apply whenever an algorithms performance is a piecewise-constant, -linear, or -- more generally -- piecewise-structured function of its parameters. Our theory also implies novel bounds for voting mechanisms and dynamic programming algorithms from computational biology.
The dialogue management component of a task-oriented dialogue system is typically optimised via reinforcement learning (RL). Optimisation via RL is highly susceptible to sample inefficiency and instability. The hierarchical approach called Feudal Dia
logue Management takes a step towards more efficient learning by decomposing the action space. However, it still suffers from instability due to the reward only being provided at the end of the dialogue. We propose the usage of an intrinsic reward based on information gain to address this issue. Our proposed reward favours actions that resolve uncertainty or query the user whenever necessary. It enables the policy to learn how to retrieve the users needs efficiently, which is an integral aspect in every task-oriented conversation. Our algorithm, which we call FeudalGain, achieves state-of-the-art results in most environments of the PyDial framework, outperforming much more complex approaches. We confirm the sample efficiency and stability of our algorithm through experiments in simulation and a human trial.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
