DREAMER: a computational framework to evaluate readiness of datasets for machine learning.

BACKGROUND: Machine learning (ML) has emerged as the predominant computational paradigm for analyzing large-scale datasets across diverse domains. The assessment of dataset quality stands as a pivotal precursor to the successful deployment of ML models. In this study, we introduce DREAMER (Data REAdiness for MachinE learning Research), an algorithmic framework leveraging supervised and unsupervised machine learning techniques to autonomously evaluate the suitability of tabular datasets for ML model development. DREAMER is openly accessible as a tool on GitHub and Docker, facilitating its adoption and further refinement within the research community.. RESULTS: The proposed model in this study was applied to three distinct tabular datasets, resulting in notable enhancements in their quality with respect to readiness for ML tasks, as assessed through established data quality metrics. Our findings demonstrate the efficacy of the framework in substantially augmenting the original dataset quality, achieved through the elimination of extraneous features and rows. This refinement yielded improved accuracy across both supervised and unsupervised learning methodologies. CONCLUSION: Our software presents an automated framework for data readiness, aimed at enhancing the integrity of raw datasets to facilitate robust utilization within ML pipelines. Through our proposed framework, we streamline the original dataset, resulting in enhanced accuracy and efficiency within the associated ML algorithms.

Ductile Effect of PGA/PCL Blending Plastics Using a Novel Ionic Chain Extender with Non-Covalent Bonds.

Polyglycolic acid (PGA) is a promising polymer in the packaging field owing to its excellent hydrolysis, heat resistance, and gas barrier properties, but it is limited in application due to its poor toughness. For this reason, a covalently bonded chain extender is introduced to increase compatibility with flexible polymers. However, covalent bonds are unfavorable for application to degradable plastics because of the energy required for reverse reactions. Therefore, we intended to effectively control the ductility of blending plastics by using a novel ionic chain extender with a relatively weaker non-covalent bond than the existing covalent bond. Polycaprolactone (PCL), which has biodegradability and flexibility, was selected as a blending polymer. For comparison, a covalently reactive chain extender (G-CE) and a non-covalently ionic chain extender (D-CE) were synthesized and compounded with blending plastics. Each chain extender improved the compatibility between PGA and PCL, and the ductility of the PGA/PCL blending plastics was more greatly enhanced with non-covalently bonded D-CE than with covalently bonded G-CE. At this time, the ductility of the PGA/PCL(90/10) blending plastic without CE was 7.2%, the ductility of blending plastic with D-CE (10D) was 26.6%, and the ductility of blending plastic with G-CE (10G) was 18.6%. Therefore, it was confirmed that the novel ionic chain extender inducing non-covalent bonds improves the compatibility between PGA and PCL and is more advantageous in enhancing ductility through a reversible reaction.

Impact of N-Substituent and pKa of Azole Rings on Fuel Cell Performance and Phosphoric Acid Loss.

The influence of N-substituent and pKa of azole rings has been investigated for the performance of high-temperature polymer electrolyte membrane fuel cells (HT-PEMFCs). Imidazole, benzimidazole, and triazole groups were functionalized on the side chains of poly(phenylene oxide), respectively. Each azole group is categorized by their N-substituent into two types: unsubstituted and methyl-substituted azoles. The membranes with methyl-substituted azoles showed higher phosphoric acid (PA) doping levels with an average increase of 20% compared to those with unsubstituted azoles in the full-doped states. However, unsubstituted azoles more effectively improved the proton conductivity and the membrane with unsubstituted imidazole (IMPPO-H) showed a high anhydrous proton conductivity of 153 mS/cm at 150 °C. In contrast, the membranes with methyl-substituted azoles showed a higher PA retention with an average increase of 81% compared to those with unsubstituted azoles. The higher PA retention of methyl-substituted azoles also led to the higher fuel cell performance with the maximum increase of 95% in the power density. It was also revealed that higher pKa of azoles enhanced the PA retention and the fuel cell performance. Based on the experimental results of PA retention and density functional theory calculations, the PA loss mechanism was also proposed.