1,720,987 research outputs found
Breast cancer with machine learning in MATLAB
Get PDFWe will look at a case study done in Wisconsin of breast cancer data and analyze various variables in order to determine whether breast tissue is malignant or benign. Through analyzing the various variables, we will use Classification Learner, Regression Leaner, and Neural Net Clustering in MATLAB, in order to determine which methods, offer the highest degree of accuracy in determining our original data. As you will see in the Classification Learner the Quadratic SVM and the Cubic SVM with thirty predictors will classify a breast cancer mass as malignant or benign with a 98.2% accuracy. And the Fine Tree algorithm with thirty predictors can classify a breast cancer mass as malignant or benign with a 93.1% accuracy. So, clearly Quadratic SVM and the Cubic SVM with thirty predictors are the better algorithms for classifying our data. And after analyzing our data in the Regression Learner depending on our response our algorithm will vary but we will still be able to determine which algorithm offers a better fit depending on model statics. And in the Neural Net Clustering app our data will be represented by just thirteen neurons due to the correlation that exists amongst our variables in our SOM Neural Net
Artificial intelligence algorithms for activity pattern detection in the information operation networks
No full textRussian Internet Trolls in Information operations use fake personas to spread disinformation through multiple social media streams. Given the increased frequency of this threat across social media platforms, understanding those operations is paramount in combating their influence. Building on existing scholarship on the inner functions within influence networks on social media, we suggest a new approach to map those types of operations. Using Twitter content identified as part of the Russian influence network, we created a predictive model to map the network operations.
We classify accounts type based on their authenticity function for a sub-sample of accounts by introducing logical categories and training a predictive model to identify similar behavior patterns across the network. Our model attains 88% prediction accuracy for the test set. Validation is done by comparing the similarities with the 3 million Russian troll tweets dataset. The result indicates a 90.7% similarity between the two datasets. Furthermore, we compare our model predictions’ on a Russian tweets dataset, and the results state that there is 90.5% correspondence between the predictions and the actual categories. The prediction and validation results suggest that our predictive model can assist with mapping the actors in such networks.
Tweet content and hashtags are the core components of Twitter. Therefore, we take advantage of investigating these essential core components to study and understand the authenticity function of actors and their behaviors. To do so, we use Natural Language Processing and cluster text data in order to map the tweets to our logical categories. Validation is performed by comparing the similarities between the logical category and the corresponding cluster. The comparison results show that the accuracy of the text data clustering can be enhanced as an independent model from the predictive model.
Visualizing and studying activities and patterns on Twitter are attractive tasks to comprehend the Russian trolls' behaviors. Due to the higher dimensional activity-related data, we use the dimensional reduction technique to visualize actions on Twitter and study their patterns. We have identified some interesting relationships and patterns, such as Tweets only, Retweets only, Tweets and Retweets only, Likes only, and More Likes and Replies.Embargo status: Restricted until 09/2027. To request the author grant access, click on the PDF link to the left
Functional Differential Equations with Piecewise Constant Argument
No full textThis book presents recent developments in nonlinear dynamics and physics with an emphasis on complex systems. The contributors provide recent theoretic developments and new techniques to solve nonlinear dynamical systems and help readers understand complexity, stochasticity, and regularity in nonlinear dynamical systems. This book covers integro-differential equation solvability, Poincare recurrences in ergodic systems, orientable horseshoe structure, analytical routes of periodic motions to chaos, grazing on impulsive differential equations, from chaos to order in coupled oscillators, and differential-invariant solutions for automorphic systems, inequality under uncertainty
Regularity and stochasticity of nonlinear dynamical systems
No full textThis book presents recent developments in nonlinear dynamics and physics with an emphasis on complex systems. The contributors provide recent theoretic developments and new techniques to solve nonlinear dynamical systems and help readers understand complexity, stochasticity, and regularity in nonlinear dynamical systems. This book covers integro-differential equation solvability, Poincare recurrences in ergodic systems, orientable horseshoe structure, analytical routes of periodic motions to chaos, grazing on impulsive differential equations, from chaos to order in coupled oscillators, and differential-invariant solutions for automorphic systems, inequality under uncertainty
Statistical Learning of Political Conflict Dynamics
Get PDFThis study aims to delve deeper into the intricacies of political violence in Africa, with a particular emphasis on providing fresh and valuable insights into this com- plex phenomenon based on empirical findings. By highlighting the importance of analyzing micro-level factors and their interrelated components when studying po- litical violence, this research aims to lay the groundwork for further investigations in this area and offers valuable insights for policymakers and decision-makers seeking to develop effective approaches to tackle this problem. The research identifies a specific time frame around elections during which the like- lihood of violent episodes substantially increases, particularly within a four-month electoral cycle. This finding underscores the importance of understanding the inter- play between political processes and violent conflicts in the African context. Fur- thermore, the findings of this study shed light on how pre- and post-election periods in Africa experience political violence in different ways, highlighting the need for detailed analysis that takes into account the specific context and timing of such con- flicts. Moreover, the research uncovers the various factors that can influence the duration of domestic extremist conflicts in the continent, thus contributing to our understand- ing of the root causes and potential solutions to political violence in Africa. Specif- ically, the study identifies the number of attacks and deaths as crucial indicators of the severity of escalation or de-escalation in rebel-state conflicts. By providing a comprehensive analysis of these factors, the research sheds light on the underlying mechanisms that drive political violence in Africa
Indirect Analysis of Public Discourse and Polarization: Structured Discussion Perception and Generational Loyalty Trends
No full textPublic discourse shapes societal opinions, influencing individual and collective perceptions.
This study examines how structured debates impact external observers, exploring
whether exposure to polarized discussions intensifies polarization or moderates
opinions. Traditional public opinion research often faces biases like social desirability
and self-censorship. To overcome these limitations, this study employs an experimental
design where observers assess debates, with responses analyzed using statistical
models, machine learning, and AI-driven semantic analysis. A key focus is the interaction
between cognitive predispositions and exposure to discussions. Findings reveal
four dimensions of loyalty—Optimism vs. Skepticism, Trust vs. Distrust, Patriotism
vs. Critical Perception, and Focus on Benefits vs. Problems—highlighting generational
shifts from collectivist to individualist values. Younger cohorts exhibit increased
skepticism, while older participants display greater institutional trust. These
insights have implications for media literacy and political communication, offering
new methodologies for understanding public sentiment and mitigating ideological polarization
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Survival under uncertainty: an introduction to probability models of social structure and evolution
No full textThis book introduces and studies a number of stochastic models of subsistence, communication, social evolution and political transition that will allow the reader to grasp the role of uncertainty as a fundamental property of our irreversible world. At the same time, it aims to bring about a more interdisciplinary and quantitative approach across very diverse fields of research in the humanities and social sciences. Through the examples treated in this work – including anthropology, demography, migration, geopolitics, management, and bioecology, among other things – evidence is gathered to show that volatile environments may change the rules of the evolutionary selection and dynamics of any social system, creating a situation of adaptive uncertainty, in particular, whenever the rate of change of the environment exceeds the rate of adaptation. Last but not least, it is hoped that this book will contribute to the understanding that inherent randomness can also be a great opportunity – for social systems and individuals alike – to help face the challenge of “survival under uncertainty”
Understanding Mexico Drug Cartel Violence Dynamics: A Comprehensive Analysis Incorporating Spatiotemporal Patterns and Socioeconomic Influences
No full textOver the past three decades, Mexico has navigated complex political, social, and economic transformations, while simultaneously grappling with the enduring challenge of drug cartel violence. This comprehensive research illuminates the intricate dynamics and determinants of Mexico’s drug cartel violence from 1989 to 2021, a period marked by significant shifts in the country’s political landscape and the escalating influence of criminal organizations. Intense and brutal violence, stemming from power struggles among various cartels have left an indelible mark on Mexican society. As the third-largest country in Latin America, Mexico’s unique geography, bordered by the United States to the north and surrounded by vast bodies of water, has presented both opportunities and challenges in addressing drug cartel violence. This research delves deeply into this multifaceted issue, employing a range of analytical techniques to examine the temporal trends, spatial distribution, relationships between actors (cartels), and key factors influencing drug cartel violence in Mexico under various political, economic, demographic, and educational backgrounds. To strengthen our analysis, we utilized data from the Uppsala Conflict Data Program Georeferenced Event Dataset (UCDP), offering a comprehensive view of drug cartel violence events in Mexico from 1989 to 2021. This dataset enabled a multifaceted analysis, including temporal trends and spatial distribution patterns of violence. Our temporal analysis, using graphical methods and the Mann-Kendall Trend Test, uncovered pivotal shifts in the conflict landscape, such as a significant surge in violence in 2007 and a resurgence in 2018. These periods signal pivotal moments in Mexico’s ongoing battle against drug cartels. The last 15 years have seen an alarming increase in violence-related casualties, highlighting the urgency of this issue. In our spatial distribution analysis, advanced mapping techniques and spatial autocorrelation assessments illuminated the geographical distribution of cartel violence. The northern border regions emerged as hotspots, with a transition from random spatial distribution to concentrated violence in specific areas from 2007 to 2021. Our study also delved into the dyadic interactions between cartels, using the Hawkes Process and extensive network analysis. The Sinaloa and Jalisco New Generation cartels were predominant, involved in over 80\% of the conflicts. Moreover, our analysis indicated that conflicts involving civilians, though rarer, are more intense, underscoring the severe impacts on civilian populations and highlighting the need for precise intervention strategies. Furthermore, the research further examined the multifaceted factors influencing drug cartel violence, such as political governance, economic conditions, demographics, and education. Notably, PRD governance correlated with the highest increase in conflicts, while PAN governance led to increases, especially at medium to high violence levels. Economic factors like personal remittances and GDP per capita significantly influenced violence patterns, while demographic and educational findings suggested complex relationships affecting conflict dynamics. In conclusion, this research offers a panoramic view of the multifaceted nature of drug cartel violence in Mexico and its determinants. These insights provide a robust foundation for policymakers, law enforcement agencies, and stakeholders to develop more effective strategies for curbing drug cartel violence and promoting stability and safety in Mexico. Our findings underscore the importance of a nuanced understanding of these dynamics for the formulation of targeted and effective interventions
- …
