Implementation of Gaussian Process Regression in Estimating Motor Vehicle Insurance Claims Reserves

This study aims to calculate the allowance for losses by applying Gaussian Process regression to estimate future claims. Modeling is done on motor vehicle insurance data. The data used in this study are historical data on PT XYZ's motor vehicle insurance business line during 2017 and 2019 (January 2017 to December 2019). Data analysis will be carried out on the 2017 - 2019 data to obtain an estimate of the claim reserves in the following year, namely 2018 - 2020. This study uses the Chain Ladder method which is the most popular loss reserving method in theory and practice. The estimation results show that the Gaussian Process Regression method is very flexible and can be applied without much adjustment. These results were also compared with the Chain Ladder method. Estimated claim reserves for PT XYZ's motor vehicle business line using the chain-ladder method, the company must provide funds for 2017 of 8,997,979,222 IDR in 2018 16,194,503,605 IDR in 2019 amounting to Rp. 1,719,764,520 for backup. Meanwhile, by using the Bayessian Gaussian Process method, the company must provide funds for 2017 of 9,060,965,077 IDR in 2018 amounting to 16,307,865,130 IDR, and in 2019 1,731,802,871 IDR for backup. The more conservative Bayessian Gaussian Process method. Motor vehicle insurance data has a short development time (claims occur) so that it is included in the short-tail type of business.

Dna Barcoding of the Ethnobotanically Important Plants of District Dera Ghazi Khan, Punjab

In the present research work, 665 plant specimens belonging to 247 species, 62 families of 27 different orders in 5 major classes of plant kingdom i.e. Mangolopsida (Dicot), Lilipopsida (Mononcot), Pinidae (Gymnosperms), Pteropsida and Bryopsida were collected from different parts of district Dera Ghazi Khan, Punjab, Pakistan. Herbarium sheets with mounted and identified specimens of all the species were deposited in Dr Sultan Ahmad herbarium, GC University, Lahore after posting voucher numbers. The ethnobotanical knowledge on the use of local plants was collected from four hundred pertinent people belonging to all walks of life through their interviews using a semi-structured questionnaire. The plants were found in use by locals as food, feed, medicine, source of fuel and furniture wood, etc. The leaves of the plants were found most frequent in use for treating various ailments, as compared to whole plant or other parts of the plant. Tissue sampling and DNA barcode sequencing was carried out at NIBGE, Faisalabad and Canadian Centre of DNA Barcoding (CCDB), University of Guelph, Canada, respectively. At Family level, 61/62 (98.39%) families were found having at least one sequence of rbcl while no matK sequence in 6/27 (22.22%) orders and 13/62 (19.67%) families. But 49/61(80.33%) families were found to have both rbcl and matK sequences. A total of 1015 barcodes were sequenced, with 630/665 (94.73%) barcodes having rbcl sequences of 235/247 (95.14%) species representing 176 genera and 62 families while 385/665 (57.90%) matK sequences of 185/247 (74.90%) species from 146 genera and 49 families. The data analysis was accomplished on the basis of three criteria; PCR Amplification and Sequencing Success of rbcl and matK; Barcode Gap Analysis (Nearest Neighbor (NN) Analysis) and Neighbor Joining (N/J) Cluster Analysis. PCR and sequencing success were found very high for the rbcl region, i.e. 94.73% (630 / 665 specimens) while matK had the lowest overall rate of recovery, i.e. 57.9% (385 / 665 specimens). The matK showed the highest sequence quality for almost all the recovered samples. Barcode Gap Analysis revealed that 95.22% sequences of matK and 96.63 % of rbcl exhibited no intra-specific variation. The inter-specific divergence varied from 0.0% to 19.06% for rbcl while 0.0% to 4.12% for matK. Out of 105 congeneric species with rbcl sequences, 69/105 (65.71%) were identified while in congeneric species with matK sequences 54/71 species (73.24%) were successfully differentiated. In pair wise divergence across all the species (non-congeneric), both xiv matK and rbcl sequences showed clear boundaries between the 84.21% (96/114) and 70.45% (93/133) of the species, respectively. In Barcode Gap Analysis, matK showed more discriminatory power than rbcl. Among taxa, patterns of sequence divergence was visualized by means of Neighbor-Joining (NJ) cluster on MEGA6. In all three trees of rbcl, matK and rbcl+matK, assignment of families within their respective orders was 83.61% (51/61), 79.59% (39/49) and 83.67% (41/49), respectively. In all three trees of rbcl, matK and rbcl+matK, species were identified on the basis of “Bootstrap Threshold Value” and “Monophyly”. In neighbor trees (N/J) of rbcl, matK and rbcl + matK, the number of monophyletic families were 49/61(80.33%), 40/49 (81.63%) and 40/47 (85.11%), respectively. In neighbor joining (N/J) trees of rbcl, matK and rbcl + matK, the number of monophyletic families were 49/61(80.33%), 40/49 (81.63%) and 40/47 (85.11%), respectively. Therefore, on the basis of monophyletic recovery and node support, all three neighbor joining trees were found best resolved monophyletic trees having more than 80% monophyletic families.