Altitude Training: Strong Bounds for Single-Layer Dropout

Authors: Stefan Wager, William Fithian, Sida Wang, Percy Liang

NeurIPS 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 6 Experiments and Discussion; Synthetic Experiment; Sentiment Classification We also examined the performance of dropout as a function of training set size on a document classification task. Figure 3a shows results on the Polarity 2.0 task [17]... We also ran experiments on a larger IMDB dataset [18]...
Researcher Affiliation Academia Stanford University, Stanford, CA-94305, USA
Pseudocode No The paper describes algorithms conceptually but does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the methodology described in this paper.
Open Datasets Yes Figure 3a shows results on the Polarity 2.0 task [17], where the goal is to classify positive versus negative movie reviews on IMDB. We also ran experiments on a larger IMDB dataset [18] with training and test sets of size 25,000 each and approximately 300,000 features.
Dataset Splits No We divided the dataset into a training set of size 1,200 and a test set of size 800; training and test sets of size 25,000 each. The paper mentions training and testing sets but does not specify a separate validation split or cross-validation strategy.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions using a “bag-of-words logistic regression model” and performing “Mc Nemar’s test” but does not provide specific ancillary software details with version numbers.
Experiment Setup No The paper mentions recalibrating the intercept and using a bag-of-words logistic regression model but does not provide specific experimental setup details such as concrete hyperparameter values, training configurations, or system-level settings in the main text.