The continuous adaptation of software systems to meet the evolving needs of users is very important for enhancing user experience (UX). User interface (UI) adaptation, which involves adjusting the layout, navigation, and content presentation based on user preferences and contextual conditions, plays an important role in achieving this goal. However, suggesting the right adaptation at the right time and in the right place remains a challenge in order to make it valuable for the end-user. To tackle this challenge, machine learning approaches could be used. In particular, we are using Reinforcement Learning (RL) due to its ability to learn from interaction with the users. In this approach, the feedback is very important and the use of physiological data could be benefitial to obtain objective insights into how users are reacting to the different adaptations. Thus, in this PhD thesis, we propose an RL-based UI adaptation framework that uses physiological data. The framework aims to learn from user interactions and make informed adaptations to improve UX. To this end, our research aims to answer the following questions: Does the use of an RL-based approach improve UX? How effective is RL in guiding UI adaptation? and Can physiological data support UI adaptation for enhancing UX? The evaluation plan involves conducting user studies to evaluate answer these questions. The empirical evaluation will provide a strong empirical foundation for building, evaluating, and improving the proposed adaptation framework. The expected contributions of this research include the development of a novel framework for intelligent Adaptive UIs, insights into the effectiveness of RL algorithms in guiding UI adaptation, the integration of physiological data as objective measures of UX, and empirical validation of the proposed framework's impact on UX.