Neural network based off-line handwritten text recognition system

This dissertation introduces a new system for handwritten text recognition based on an improved neural network design. Most of the existing neural networks treat mean square error function as the standard error function. The system as proposed in this dissertation utilizes the mean quartic error function, where the third and fourth derivatives are non-zero. Consequently, many improvements on the training methods were achieved. The training results are carefully assessed before and after the update. To evaluate the performance of a training system, there are three essential factors to be considered, and they are from high to low importance priority: (1) error rate on testing set, (2) processing time needed to recognize a segmented character and (3) the total training time and subsequently the total testing time. It is observed that bounded training methods accelerate the training process, while semi-third order training methods, next-minimal training methods, and preprocessing operations reduce the error rate on the testing set. Empirical observations suggest that two combinations of training methods are needed for different case character recognition. Since character segmentation is required for word and sentence recognition, this dissertation provides also an effective rule-based segmentation method, which is different from the conventional adaptive segmentation methods. Dictionary-based correction is utilized to correct mistakes resulting from the recognition and segmentation phases. The integration of the segmentation methods with the handwritten character recognition algorithm yielded an accuracy of 92% for lower case characters and 97% for upper case characters. In the testing phase, the database consists of 20,000 handwritten characters, with 10,000 for each case. The testing phase on the recognition 10,000 handwritten characters required 8.5 seconds in processing time.

Effects of sex, third grade reading achievement and motivation as predictors of fourth grade reading achievement of Hispanic students: A path analysis

This study explored the topic of motivation for intermediate students combining both an objective criterion measure (i.e., standardized test scores) and the self-report of students on self-concept and value of reading. The purpose of this study was to examine how third grade reading achievement correlated with the motivation of fourth grade boys and girls, and, in turn, how motivation related to fourth grade reading achievement. The participants were fourth grade students (n=207) attending two public, elementary schools in Miami-Dade County who were of primarily Hispanic origin or descent. Data were collected using the Reading Survey portion of the Motivation to Read Profile (1996) which measures self-concept and value of reading in order to measure motivation and the Third and Fourth Grade Reading Florida Comprehensive Assessment Tests 2.0 (FCAT 2.0) to assess achievement. First, a one way Analysis of Variance (ANOVA) was conducted to determine whether motivation differed significantly between fourth grade boys and girls. Second, a path analysis was used to determine whether motivation mediated or moderated the association between FCAT 2.0 third and fourth grade scores. Results of the ANOVA indicated that motivation, as measured by the Motivation to Read Profile did not differ significantly by sex. Results from the path analysis indicated that the model was significant and that third grade FCAT 2.0 scores accounted for a significant amount of the variance in fourth grade FCAT 2.0 scores once motivation was entered. Results of the study demonstrated that motivation partially mediates, but does not moderate the relationship between FCAT 2.0 third and fourth grade scores. In conclusion, it can be determined that past student achievement for fourth grade students plays a role in current student achievement when motivation is also considered. It is therefore important in order to improve the quality of fourth grade student's current performance to take into account a student's motivation and past achievement. An effort must be made to address students' motivational needs whether through school wide programs or at the classroom level in addition or in conjunction with cognition. Future research on the effect of self-concept in reading achievement is recommended.