A Combined Deep-Learning and Transfer-Learning Approach for Supporting Social Influence Prediction

Cuzzocrea, Alfredo; Leung, Carson K.; Deng, Dayong; Jason, Jiaxing; Jiang, Fan; Fadda, Edoardo

doi:10.1016/j.procs.2020.10.025

Cited by 13 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even more difficult to estimate than the previous quantities is the dynamic of the influence. In fact, this is a broad and open area of research (see, e.g., [14]). Since the proposed techniques are agnostic with respect to the dynamic of influence chosen, we test them against the models presented in Section 3, where the parameters for the CTM are reported in Table 2.…”

Section: Experimental Settingmentioning

confidence: 99%

Math‐based reinforcement learning for the adaptive budgeted influence maximization problem

Fadda,

Corso,

Brusco

et al. 2023

Networks

Self Cite

View full text Add to dashboard Cite

In social networks, the influence maximization problem requires selecting an initial set of nodes to influence so that the spread of influence can reach its maximum under certain diffusion models. Usually, the problem is formulated in a two‐stage un‐budgeted fashion: The decision maker selects a given number of nodes to influence and observes the results. In the adaptive version of the problem, it is possible to select the nodes at each time step of a given time interval. This allows the decision‐maker to exploit the observation of the propagation and to make better decisions. This paper considers the adaptive budgeted influence maximization problem, that is, the adaptive problem in which the decision maker has a finite budget to influence the nodes, and each node requires a cost to be influenced. We present two solution techniques: The first is an approximated value iteration leveraging mixed integer linear problems while the second exploits new concepts from graph neural networks. Extensive numerical experiments demonstrate the effectiveness of the proposed approaches.

show abstract

Section: Experimental Settingmentioning

confidence: 99%

Math‐based reinforcement learning for the adaptive budgeted influence maximization problem

Fadda,

Corso,

Brusco

et al. 2023

Networks

Self Cite

View full text Add to dashboard Cite

show abstract

“…Because the improved PageRank algorithm is not sensitive to the initial value of the evaluation object, the initial value of PageRank of each team is assigned as 1 and calculated by formula (3). e matrix reaches a stable state at the 7th iteration, and the team's PR value does not change any more.…”

Section: Data Selection and Analysismentioning

confidence: 99%

“…Since the end of the 20th century, a large number of researchers began to use Machine Learning Algorithm to predict the results. Cuzzocrea et al combined the deep-learning and transfer-learning approach for supporting social influence prediction [3]. Huang et al and Liu and Zhu predicted different target domains based on Pag-eRank and HITS algorithm [4,5].…”

Section: Introductionmentioning

confidence: 99%

The Ranking Prediction of NBA Playoffs Based on Improved PageRank Algorithm

Yang

Zhang²

2021

Complexity

View full text Add to dashboard Cite

It is of great significance to predict the results accurately based on the statistics of sports competition for participants research, commercial cooperation, advertising, and gambling profit. Aiming at the phenomenon that the PageRank page sorting algorithm is prone to subject deviation, the category similarity between pages is introduced into the PageRank algorithm. In the PR value calculation formula of the PageRank algorithm, the factor W(u, v) between pages is added to replace the original Nu (the number of links to page u). In this way, the content category between pages is considered, and the shortcoming of theme deviation will be improved. The time feedback factor in the PageRank-time algorithm is used for reference, and the time feedback factor is added to the first improved PR value calculation formula. Based on statistics from 1230 games during the NBA 2018-2019 regular season, this paper ranks the team strength with improved PageRank algorithm and compares the results with the ranking of regular-season points and the result of playoffs. The results show that it is consistent with the regular-season points ranking in the eastern division by the use of improved PageRank algorithm, but there is a difference in the second ranking in the western division. In the prediction of top four in playoffs, it predicts three of the four teams.

show abstract

Financial Forecasting via Deep-Learning and Machine-Learning Tools over Two-Dimensional Objects Transformed from Time Series

Baldo¹,

Cuzzocrea

Fadda

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

A Combined Deep-Learning and Transfer-Learning Approach for Supporting Social Influence Prediction

Cited by 13 publications

References 9 publications

Math‐based reinforcement learning for the adaptive budgeted influence maximization problem

Math‐based reinforcement learning for the adaptive budgeted influence maximization problem

The Ranking Prediction of NBA Playoffs Based on Improved PageRank Algorithm

Financial Forecasting via Deep-Learning and Machine-Learning Tools over Two-Dimensional Objects Transformed from Time Series

Contact Info

Product

Resources

About