Whether the Support Region of Three-Bit Uniform Quantizer Has a Strong Impact on Post-Training Quantization for MNIST Dataset?

Driven by the need for the compression of weights in neural networks (NNs), which is especially beneficial for edge devices with a constrained resource, and by the need to utilize the simplest possible quantization model, in this paper, we study the performance of three-bit post-training uniform quantization. The goal is to put various choices of the key parameter of the quantizer in question (support region threshold) in one place and provide a detailed overview of this choice's impact on the performance of post-training quantization for the MNIST dataset. Specifically, we analyze whether it is possible to preserve the accuracy of the two NN models (MLP and CNN) to a great extent with the very simple three-bit uniform quantizer, regardless of the choice of the key parameter. Moreover, our goal is to answer the question of whether it is of the utmost importance in post-training three-bit uniform quantization, as it is in quantization, to determine the optimal support region threshold value of the quantizer to achieve some predefined accuracy of the quantized neural network (QNN). The results show that the choice of the support region threshold value of the three-bit uniform quantizer does not have such a strong impact on the accuracy of the QNNs, which is not the case with two-bit uniform post-training quantization, when applied in MLP for the same classification task. Accordingly, one can anticipate that due to this special property, the post-training quantization model in question can be greatly exploited.

Demographic risk factors for suicide and depression among Serbian medical school students.

PURPOSE: The purpose of the study was to identify demographic risk factors associated with depression and suicidality among medical school students in Serbia, as well as to determine the prevalence of symptoms of depression and suicide attempts in the student population. METHOD: This was a cross-sectional study conducted on the sample of 1,296 students recruited from the School of Medicine, University of Belgrade. Following questionnaires were assigned to participating subjects: PHQ-9, Suicide Behaviors Questionnaire, and Social-demographic survey. RESULTS: PHQ-9 scores in our sample ranged from 0 to 27, averaging at 6.21 (SD = 4.96). Symptoms of depression were associated with grade point average, age, relationship status, and gender. Reported history of suicide attempts was associated with PHQ-9 scores, grade point average, relationship status, gender, history of mental illness and drug use. CONCLUSION: Our findings suggest that medical school students in Serbia may be considered an "at-risk" population. Preventative programs in the future should target symptoms of depression and should provide students with adequate mechanisms to cope successfully with school-related pressure. Alternative stress management strategies may also be useful, especially for younger students who seem to be at greater risk for depression and suicide.

Self-medication practices and risk factors for self-medication among medical students in Belgrade, Serbia.

INTRODUCTION: Self-medication among future health care professionals can represent a serious threat to professionalism in medicine and it has potential to put at risk public trust into this profession. The aim of this research was to investigate prevalence and risk factors for self-medication among population of medical students, because it was previously shown that their attitudes towards pharmacotherapy could affect the way they could prescribe medication in the future. MATERIAL AND METHODS: Research was performed as a cross-sectional study and it included 1296 (84.1%) 1st, 3rd and 6th year students of School of Medicine, University of Belgrade. Students filled out a demographic and self-medication questionnaire created for the purpose of this research and the Physical Health Questionnaire - 9 (PHQ-9). Questions about self-medication were related to the period of the previous year. RESULTS: Self-medication was reported by 79.9% students. The most frequently self-prescribed medications were analgesics (55.4%). Independent risk factors for self-medication were possession of home-pharmacies (OR = 5.3, CI 95% 3.89-7.23), lower level of father's education (OR = 1.6, CI 95% 1.18-2.25), consumption of alcoholic beverages (OR = 1.5, CI 95% 1.13-2.08), less than 1 hour spent in physical activity per week (OR = 1.4, CI 95% 1.00-2.02), female gender (OR = 1.4, CI 95% 1.02-1.89), older age (OR = 1.1, CI 95% 1.07-1.21) and higher PHQ-9 score (OR = 1.09, CI 95% 1.05-1.12). CONCLUSIONS: Self-medication is an important issue among population of medical students. Prevalence of self-medication could be controlled through regulatory authorities and further education.