Disguise Adversarial Networks for Click-through Rate Prediction
Authors: Yue Deng, Yilin Shen, Hongxia Jin
IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We applied DAN to two Ads datasets including both mobile and display Ads for CTR prediction. The results showed that our DAN approach significantly outperformed other supervised learning and generative adversarial networks (GAN) in CTR prediction. |
| Researcher Affiliation | Industry | Yue Deng, Yilin Shen, Hongxia Jin Samsung Research America, Mountain View, CA, USA {y1.deng, yilin.shen, hongxia.jin}@samsung.com |
| Pseudocode | Yes | Algorithm 1: Training DAN |
| Open Source Code | No | The paper does not provide any links to open-source code or state that code is available. |
| Open Datasets | Yes | [Ava, 2015] Avazu mobile ads ctr dataset. https://www.kaggle.com/c/avazu-ctr-prediction/data, 2015. [Cri, 2015] Criteo display ads ctr dataset. https://www.kaggle.com/c/criteo-display-ad-challenge, 2015. |
| Dataset Splits | Yes | We obey the time order and uniformly divide each dataset as 100 bulks. Each bulk contains 1% Ads impressions in a certain period and different bulks are consecutive in time. ...we train our model with the last 20 bulks of Ads impressions in the history and predict the CTR in the next 5 bulks. |
| Hardware Specification | No | Our practical training always requires 3 hours to finish 25 epochs on 9 million historic data with 4 GPUs parallelized. (The paper mentions '4 GPUs' but does not specify the model or other hardware details.) |
| Software Dependencies | No | The paper mentions implementing neural networks with Multi-Layer Perceptrons (MLP) but does not specify any software names with version numbers (e.g., Python, TensorFlow, PyTorch versions). |
| Experiment Setup | Yes | In detail, both the disguise and discriminator neural networks are configured with 4 layers and each layer contains 32 nodes. The output layer of the disguise neural network shares the same nodes number as its input layer. The output of the discriminator neural network is a sigmoid function indicating the clicking probability. We train both the DAN and the discriminator neural network for 25 epochs. |