Several studies have investigated the need for learning difficulties identification specifically Dyslexia, Dysgraphia and Dyscalculia. The identification of these difficulties among children is a multiple screening process under psychologist's supervision. Learning difficulties identification is a difficult task; it affects the learning process and the academic achievements of a child. The introduction of an Intelligent Tutoring System (ITS) to identify learning problems and teach the learning disabled through ITS is an unexplored domain. An ITS in education is extensively considered for the teaching and learning process as it is an adaptive and learner specific computer system. The capabilities of an ITS in integration with AI methodologies have put together promising results. The ITS framework implemented in this study is developed for learning disabilities identification and we have assessed total 24 participants (with or without Learning Disabilities) for the experiment. This ITS framework design is based on a pretest analysis through initial screening and then system based screening of a child response for Learning Difficulties (LDs) identification. The system based screening is implemented using neural network classifiers to identify learning difficulties. The fuzzy min-max neural network (FMNN) classification is applied to determine learner profile, learning disabled, and present learner-centered content. Fuzzy sets as pattern classes are introduced in supervised learning neural network classification for learner profiling of learning Disabled in an ITS. The results are generated based on the classification applied to the input provided during the pre-test. The results indicate that the integration of fuzzy with the neural network has significantly increased the ITS accuracy.