Visual Learning of Arithmetic Operation

Authors: Yedid Hoshen, Shmuel Peleg

AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments were carried out; The correctness of the test set was measured using an OCR software; The effectiveness of the neural network approach has been tested on the following operations.
Researcher Affiliation Academia Yedid Hoshen and Shmuel Peleg School of Computer Science and Engineering The Hebrew University of Jerusalem Jerusalem, Israel
Pseudocode No The paper describes the network operation and equations in text and diagrams (Figure 6) but does not include any explicit pseudocode blocks or algorithms.
Open Source Code No The paper states 'The network was implemented using the Caffe package (Jia et al. 2014)' which refers to a third-party tool, but it does not provide a link or explicit statement about the availability of their own source code for the described methodology.
Open Datasets No The input integers are randomly selected in a prespecified range (for addition we use the range of [0,4999999]), and are written on the input pictures... Learning consists of training the network with N such input/output examples (we use N = 150, 000). The paper describes generating its own dataset but does not provide specific access information for public availability.
Dataset Splits No The paper states '150,000 input/output pairs were randomly generated for training and 30,000 pairs were randomly created for testing,' but no specific details about a separate validation set or a three-way split for hyperparameter tuning were provided.
Hardware Specification No The paper does not provide specific hardware details such as GPU or CPU models used for running the experiments.
Software Dependencies No The paper mentions 'The network was implemented using the Caffe package (Jia et al. 2014)' but does not specify a version number for Caffe or any other software dependencies.
Experiment Setup Yes The network has three hidden layers each with 256 nodes with Re LU activation functions... and an output layer... with sigmoid activation. All nodes between adjacent layers were fully connected. An L2 loss function is used... The network is trained via minibatch stochastic gradient descent with learning rate 0.1, momentum 0.9 and mini-batch size was 256. 50 epochs of training were carried out.