The field of influencer marketing has experienced rapid growth in recent years. However, uncovering the true effectiveness of this marketing approach remains a significant challenge. This thesis addresses the challenge of predicting the effectiveness of influencer marketing campaigns by employing advanced machine learning techniques, specifically the Auto Machine Learning framework Autogluon. With the aim of democratizing machine learning and empowering businesses in the influencer marketing domain, this work leverages Autogluon to predict the interactions generated by influencers when posting affiliate links. By evaluating various settings of AutoGluon and assessing the performance using metrics such as R-squared score, we observed promising results with good predictive accuracy. The findings from our study contribute to critical discussions in the field. This research offers a streamlined and efficient approach to machine learning, reducing the need for extensive manual model tuning and enabling marketers to make informed decisions and optimize their campaign strategies. The outcomes of this study have practical implications for businesses, allowing them to effectively predict campaign interactions and maximize the impact of influencer marketing initiatives. By leveraging the power of automated machine learning, this thesis opens up new opportunities for businesses to harness the potential of influencer marketing in driving successful marketing campaigns. / Influencer marketing trenden har ökat markant de senaste aren men effektiviteten av denna marknadsföringsmetod är till stor del oviss. Denna avhandling utfårskar svårigheten med att förutse effekten av influencer marketing kampanjer med hjälp av avancerad maskininlärningsteknik, specifikt Auto Machine Learning-ramverket Autogluon. Med målet att demokratiserar och uppmuntra företag att använda maskininlärning, utforskar denna avhandling Autogluon för att förutse interageringar som genereras när influerare publicerar affiliate länkar. Genom att utvärdera olika inställningar av Autogluon och analysera olika data som till exempel R-kvadratvärde observerade vi lovande resultat med god förutsägbar precision. Resultaten från vår studie bidrar till kritiska diskussioner inom området. Denna forskning erbjuder en strömlinjeformad och effektiv metod för maskininlärning, vilket minskar behovet av omfattande manuellt modellarbete och möjliggör för marknadsförare att fatta informerade beslut och optimera sina kampanjstrategier. Resultaten av denna studie har praktiska implikationer för företag, vilket gör det möjligt för dem att effektivt förutse interaktioner i kampanjer och maximera effekten av influencer marketingvertyg. Genom att applicera automatiserad maskininlärning öppnar denna avhandling nya möjligheter för företag att dra nytta av potentialen hos influencer marketing för att driva framgångsrika marknadsföringskampanjer.

This master thesis presents a machine learning framework for real-time gesture and skeleton-based action recognition, integrated with the Unity game engine. The system aims to enhance human-computer interaction (HCI) in gaming and 3D related applications through natural movement recognition, by training a model on skeleton tracking data. The framework is trained to accurately categorize and identify gestures such as kicks and punches, enabling a more immersive gaming experience not existing in traditional controllers. After studying the evolution of HCI and how machine learning has transformed and reshaped the interaction paradigm, the prototype system is built through data collection, augmenting, and preprocessing, followed by training and evaluating a Long Short-Term Memory (LSTM) neural network model for gesture classification. The model is integrated into Unity via Unity Sentis using Open Neural Network Exchange (ONNX) format, enabling efficient real-time action recognition in 3D space. Each component of the pipeline is available and adaptable for future custom- ization and needs, skeleton tracking and Unity integration is built using the ZED 2i camera and ZED SDK. Experimental results demonstrate that the system presented can achieve over 90% accuracy in identifying predefined gestures. As a bridging solution tailored for Unity, this framework offers a practical solution to action recognition that could be found useful in future applications. This work contributes to advancing human-computer interaction and offers a foundation for further development in gesture-based Unity game design.