REFORMS: Consensus-based Recommendations for Machine-learning-based Science.

Machine learning (ML) methods are proliferating in scientific research. However, the adoption of these methods has been accompanied by failures of validity, reproducibility, and generalizability. These failures can hinder scientific progress, lead to false consensus around invalid claims, and undermine the credibility of ML-based science. ML methods are often applied and fail in similar ways across disciplines. Motivated by this observation, our goal is to provide clear recommendations for conducting and reporting ML-based science. Drawing from an extensive review of past literature, we present the REFORMS checklist (recommendations for machine-learning-based science). It consists of 32 questions and a paired set of guidelines. REFORMS was developed on the basis of a consensus of 19 researchers across computer science, data science, mathematics, social sciences, and biomedical sciences. REFORMS can serve as a resource for researchers when designing and implementing a study, for referees when reviewing papers, and for journals when enforcing standards for transparency and reproducibility.

Performance of vertical up-flow-constructed wetlands integrating with microbial fuel cell (VFCW-MFC) treating ammonium in domestic wastewater.

Vertical up-flow-constructed wetlands integrating with microbial fuel cell (VFCW-MFC) were evaluated for NH4+-N removal and bioelectricity recovery. The experiments were carried out in lab-scale VFCW-MFC microcosms treating synthetic domestic wastewater under different operational conditions of pH, hydraulic retention time, and mass loading rate. Effects of wild ornamental grass (Cenchrus setaceus) on treatment performance and voltage output were investigated simultaneously. Experiments demonstrated that the neutral pH of influents favoured NH4+-N removal and power generation. Extended retention time improved treatment capacity and power output but likely depended on the substrate availability. COD removal and power output increased, while NH4+-N removal decreased with increasing mass loading rates. At the loading rate of 88.31 mg COD/L.day, planted VFCW-MFCs exhibited better NH4+-N treatment performance (36.9%) and higher voltage output (132%-143%) than unplanted systems. Plants did not affect the COD removal efficiency of VFCW-MFCs (>95%). Power density was in the range of 1.26-1.59 mW/m2 in planted microcosms with a maximum CE of 13.6%. The anode layer accounted for a major proportion of NH4+-N removal in VFCW-MFCs. This study implies that NH4+-N in domestic wastewaters with relatively high COD:N ratios can be treated effectively in up-flow CW-MFCs via anaerobic processes, including anammox and heterotrophic denitrifying processes. The mass loading rate could be a critical parameter to balance different microbial processes, thus, coincidently determining the potential of power recovery from wastewaters.