An empirical analysis of homicides in Mexico through Machine Learning and statistical design of experiments

Homicide is one of the most important mortality causes that has reduced the Mexican life expectancy. That is why the aim of this work is to identify some sociodemographic and economic factors that can help explain homicides in Mexico and measure their impact, assuming the current conditions prevail. To do that, several Machine Learning (ML) methods were evaluated. The C5.0 model is best suited for the data at hand. After fine-tuning the algorithm, we used the estimated model to identify the main factors that explain homicides. Among these factors, eleven were selected that can be influenced by direct changes in domestic public policy, laws and/or regulations. These were used as input in a two-level fractional factorial Statistical Design of Experiments (DOE) to estimate their main effects and possible interactions. Although several of these factors had statistically significant effects on homicide rate, the one that had the biggest and direct impact from a practical perspective, was the Rule of Law Index (RLI). In fact, if we assumed that all states had the median RLI of 0.37, implementing domestic policies and procedures to move them all to the best RLI level could significantly reduce homicide rates.

El homicidio es una de las principales causas de muerte que ha reducido la esperanza de vida de los mexicanos. El objetivo de este trabajo es identificar algunos factores sociodemográficos y económicos que puedan ayudar a explicar homicidios en México y medir su impacto, suponiendo que las condiciones actuales permanecen. Para lograrlo, comparamos diferentes métodos de Aprendizaje de Máquina (AM). Para tal fin, se encuentra que el modelo C5.0 es el más adecuado. Después de hacer una calibración final del modelo, lo utilizamos para determinar los veinticinco principales factores que explican el fenómeno de homicidios. Se seleccionan 11 factores que se consideran pueden ser influenciados directamente por cambios en políticas públicas, leyes y/o regulaciones. Estos predictores fueron utilizados como entrada en un diseño de experimentos factorial fraccionado con dos niveles para estimar los principales efectos principales e interacciones posibles. A pesar de que varios de estos factores tuvieron impactos estadísticamente significativos, el que mostró tener el mayor impacto directo desde una perspectiva práctica fue el Índice de Estado de Derecho (IED). De hecho, asumiendo que todos los estados tuvieran el valor de IED de 0.37, correspondiente a la mediana en todo el país, si se implementaran políticas y procedimientos para ubicar a todos los estados al nivel del mejor estado en términos de IED, se lograría una reducción altamente significativa en la incidencia de homicidios en México.

Bacillus licheniformis SMIA-2 protease used as an additive in detergent formulations

Microbial proteases, especially from Bacillus spp., have been widely exploited for a broad variety of applications, such as the improvement of the cleaning efficiency of conventional detergents. In this work, the statistical design of the experiment was used to optimize the concentrations of a three-component mixture: Bacillus licheniformis SMIA-2 protease, Linear Alkylbenzene Sulphate and hydrogen peroxide, in an attempt to prepare an environmentally correct cleaning formulation. The results demonstrated that the combination of 1% (w/v) protease with 1.5% (w/v) LAS and 1% (w/v) H2O2 was effective in removing blood from cloth pieces and that a protease concentration decrease from 1.0% to 0.5% (w/v) would not have a signi&64257;cant impact on percent blood removal if LAS concentrations between 1.5-2.0% (w/v) in combination with lower (<0.5%, w/v) concentrations of H2O2 were used. Thus, the protease from Bacillus licheniformis SMIA-2 can be effectively incorporated into cleaning formulations together with LAS and hydrogen peroxide to formulate more sustainable detergents.

Optimization and quality evaluation of ginsenoside Rg3 self-emulsifying drug delivery by Box-Behnken experimental design / 中草药

Objective: To optimize the formulation of ginsenoside Rg3 self-emulsifying drug delivery by Box-Behnken experimental design (BBD) and evaluate its quality. Methods: Formulations were optimized using design of experiments by employing a 3-factor, 3-level BBD. Independent variables studied were the emulsifier, co-emulsifier, and the oil phase. The dependent variables were drug loading and average particle size. Through the BBD software, the response surface plots were drawn and the polynomial of influence of each factor index was obtained. Finally the formula optimization was obtained and the efficiency, droplet size, and dissolution rate of self-emulsifying drug delivery were investigated. Results: Optimum prescription was as follows: emulsifier [RH-400-Triton X-100 (1∶1)], co-emulsifier (isopropanol), and the oil phase (ethyl oleate). The optimum proportion was 3∶1.475∶0.5. The optimized ginsenoside Rg3 self-emulsifying drug delivery could fully emulsify in 5 min and the average drug loading was 917.53 μg/mL, average particle size was 111.6 nm. The dissolution degree in vitro of the formulations showed that the release rate and level were apparently superior to the conventional capsules. Conclusion: The observed responses are in close agreement with the predicted values of the mathematic models, and the BBD is suitable for optimizing ginsenoside Rg3 self-microemulsifying drug delivery.

Effect of Process Variables on Survival of Bacteria in Probiotics Enriched Pomegranate Juice.

Aims: The consumption of foods and beverages containing probiotic microorganisms is a growing, global consumer trend. In this research, production of probiotic pomegranate juice containing Lactobacillus plantarum and Lactobacillus delbrueckii was studied. Study Design: Plackett-Burman statistical design was used to evaluate the impact of eleven process variables on the viability of both probiotics. Impact of incorporation of grape juice, tomato juice and pomegranate peel extract as well as phenolic compounds and vitamins have been investigated. Place and Duration of Study: Department of Food Science and Technology, Varamin Branch, Islamic Azad University, Varamin, Tehran, Iran, between Sep 2012 and July 2013. Methodology: Pomegranate juices were inoculated with probiotic bacteria and their survival was evaluated every week by pure plate method. The effect of 11 variables (in two levels) on survival of bacteria by the statistical design of Plackett-Burman was evaluated. For this purpose, 12 treatments in triplicate by the Minitab (version = 11.0) software at significant levels α= .01 were analyzed. Results: The highest survival rate of L. plantarum (4.74 ×106 CFU/mL) and L. delbrueckii (4 ×106 CFU/mL) was obtained by 10% v/v inoculation of a 48 h inoculum culture in MRS broth medium to enriched pomegranate juice (10% v/v Grape juice, 5% v/v tomato juice, 0.1% v/v pomegranate peel extract and 2.0 g.L glucose) which was inoculated in anaerobic condition for 72 h at 37ºC and kept for 2 weeks at environment temperature. Sensory evaluation shows the probiotic juice was accepted by consumers with no significant difference in comparison to control in terms of taste, odour and overall acceptability (P>.05). Conclusion: The results of this study suggest that grape, tomato juices and pomegranate peel extract exert a protective effect on L. plantarum and L. delbrueckii viability under acidic condition of pomegranate juice and storage time, which was associated with the chemical composition of them. This study indicates that develop of probiotic pomegranate juices with acceptable viability and stability of the probiotic is possible.

Fabrication and Optimization of Novel Lornoxicam Matrix Tablets Using 3-Factor 3-Level Box-Behnken Statistical Design: Invitro and Invivo Evaluation.

In the present study efforts have been made to prepare sustained release matrix tablets of Lornoxicam. Matrix tablets were prepared by direct compression method by using Hydroxypropyl methyl cellulose K15 (HPMC- K15), Ethyl cellulose (EC) and Sodium carboxy methyl cellulose (Na-CMC) as polymers in different concentrations. A 3-factor 3- level Box-Behnken statistical design was used as an optimization tool having total of 17 experimental runs with 5 central points. All three polymers were selected as independent variables while %age drug release at various time intervals and hardness were used as dependant variables. In vivo studies were conducted on human plasma using Tenoxicam as internal standered. All the detections were made on SYKNM HPLC. Foriour Transform Infrared Spectroscopy (FTIR) and Differential Scanning Calorimetery (DSC) studies were conducted and no chemical interaction was found between drug and polymers. The drug release mechanism was mainly governed by non-fickian (anomalous) diffusion and zero-order (case II) transport diffusion. Regression analysis was performed on dissolution data obtained with the selected response variables and polynomial models were constructed. Polynomial models were further validated using one way ANOVA and results indicated that all the polymers used have significant effect on selected response (p>0.05). Contour plots and three dimensional response surface curves were drawn. In- vivo studies were conducted on two tablet formulation indicating slow and sustained release of the drug from matrix. From Behnken design it is possible to successfully formulate and optimize Lornoxicam sustained release matrix tablets with three polymers (HPMC- K15, EC and Na-CMC) in combination.