Effects of Random Seeds on the Accuracy of Convolutional Neural Networks

Fellicious, Christofer; Weißgerber, Thomas

doi:10.1007/978-3-030-64580-9_8

Cited by 6 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4.1, by using the original source code and data. This experiment has two goals: (i) verify that published results are replicable; and (ii) measure the numerical stability of CFGAN given the stochastic nature of its architecture [9,25].…”

Section: Rq1: Cfgan Replicability and Numerical Stabilitymentioning

confidence: 98%

“…This discussion is based on the findings of [23], which highlights the importance of describing not only how a model works, but also what works and why it works, as well as how experimental inquiries that aim to deepen our understanding are valuable research contributions even when no new algorithm is proposed. Furthermore, this work analyzes the replicability, reproducibility, and recommendation quality of CFGAN [4] as well as its numerical stability which is known to be a challenge for GANs [9,25]. The main research questions of this work are: RQ1: Is CFGAN replicable and numerically stable?…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An Evaluation Study of Generative Adversarial Networks for Collaborative Filtering

Maurera

Dacrema

Cremonesi

2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

This work explores the reproducibility of CFGAN. CFGAN and its family of models (TagRec, MTPR, and CRGAN) learn to generate personalized and fake-but-realistic rankings of preferences for top-N recommendations by using previous interactions. This work successfully replicates the results published in the original paper and discusses the impact of certain differences between the CFGAN framework and the model used in the original evaluation. The absence of random noise and the use of real user profiles as condition vectors leaves the generator prone to learn a degenerate solution in which the output vector is identical to the input vector, therefore, behaving essentially as a simple autoencoder. The work further expands the experimental analysis comparing CFGAN against a selection of simple and well-known properly optimized baselines, observing that CFGAN is not consistently competitive against them despite its high computational cost. To ensure the reproducibility of these analyses, this work describes the experimental methodology and publishes all datasets and source code.

show abstract

Section: Rq1: Cfgan Replicability and Numerical Stabilitymentioning

confidence: 98%