Phoenix dactylifera L., usually identified as the Date Palm, is defined as a palm tree in the genus Phoenix, known for its comestible crop. To facilitate characterization of date-palm germplasm, a model based on morphological parameters has been designed to explain the effect of these descriptors on development of different date-palm accessions located in the Tunisian eco-geographical environment. In our study, we relate the phenotypic differences and connections in a collection of 69 various date-palm accessions as showed by vegetative factors. Furthermore, we apply morphological markers to describe date-palm ecotypes. Morphological and phenotypical characteristics depicting the vegetative models were estimated. The measured data set was treated using principal components analysis and principal coordinate analysis clustering. The principal component analysis indicated that the first two principal components represented 39% of the totality of variation. Indeed, the Principal coordinate analysis plot demonstrated that first three coordinates explained more than 43% of the total variance. Obtained results showed that shaft length increased with spine length and stipe size has inverse relationship with crown shape. Besides, spine rigidity has a growing relationship with leaflet width and palm bunch size has an increasing relationship with both leaflet length and leaf palm length. In addition, Bayesian networks gave an explicit model describing the different relationships between fruit and kernel characteristics. Decision-tree technique demonstrated that fruit and kernel characteristics (fruit and kernel weights, fruit and kernel lengths, fruit and kernel widths) were indispensable for model interpretation and thus palm-date yield prediction. This study supplies a simple classification plan and comprehensible analytical devices for recognizing the most important elements combined that may be practical to develop programs for date collect management.