Feature Noise Resilient for QoS Prediction with Probabilistic Deep Supervision

TL;DR

The Probabilistic Deep Supervision Network (PDS-Net) is presented, a robust framework designed to effectively identify and mitigate feature noise, thereby improving QoS prediction accuracy and highlighting its robustness and versatility across diverse QoS prediction environments.

Abstract

Accurate Quality of Service (QoS) prediction is essential for enhancing user satisfaction in web recommendation systems, yet existing prediction models often overlook feature noise, focusing predominantly on label noise. In this paper, we present the Probabilistic Deep Supervision Network (PDS-Net), a robust framework designed to effectively identify and mitigate feature noise, thereby improving QoS prediction accuracy. PDS-Net operates with a dual-branch architecture: the main branch utilizes a decoder network to learn a Gaussian-based prior distribution from known features, while the second branch derives a posterior distribution based on true labels. A key innovation of PDS-Net is its condition-based noise recognition loss function, which enables precise identification of noisy features in objects (users or services). Once noisy features are identified, PDS-Net refines the feature's prior distribution, aligning it with the posterior distribution, and propagates this adjusted distribution to intermediate layers, effectively reducing noise interference. Extensive experiments conducted on two real-world QoS datasets demonstrate that PDS-Net consistently outperforms existing models, achieving an average improvement of 8.91% in MAE on Dataset D1 and 8.32% on Dataset D2 compared to the ate-of-the-art. These results highlight PDS-Net's ability to accurately capture complex user-service relationships and handle feature noise, underscoring its robustness and versatility across diverse QoS prediction environments.

Figures (9)

• Figure 1: The framework of PDS-Net. Prior net is a decoder network; $P_1$ is a Gaussian-based prior distribution learned from known information; $P_2$ and $P_i$ are priori distributions based on deep variable in the middle layer. The posterior distribution $P_r$ is learned from real label.
• Figure 2: QoS distribution of datasets
• Figure 3: Probability distribution learning process.$P_1$ is the probability distribution learned from the missing features. $P_r$ is a probability distribution learned from ground truth labels. $P_1'$ is the distribution of p1 after training through $P_r$. $P_2$ is the probability distribution of deep features in the middle layer deep probabilistic supervision
• Figure 4: The impact of probabilistic supervision module
• Figure 5: The results of PDS-Net with different cond in loss