Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society
Fairness in machine learning is crucial when individuals are subject to automated decisions made ... more Fairness in machine learning is crucial when individuals are subject to automated decisions made by models in high-stake domains. Organizations that employ these models may also need to satisfy regulations that promote responsible and ethical A.I. While fairness metrics relying on comparing model error rates across subpopulations have been widely investigated for the detection and mitigation of bias, fairness in terms of the equalized ability to achieve recourse for different protected attribute groups has been relatively unexplored. We present a novel formulation for training neural networks that considers the distance of data points to the decision boundary such that the new objective: (1) reduces the average distance to the decision boundary between two groups for individuals subject to a negative outcome in each group, i.e. the network is more fair with respect to the ability to obtain recourse, and (2) increases the average distance of data points to the boundary to promote adversarial robustness. We demonstrate that training with this loss yields more fair and robust neural networks with similar accuracies to models trained without it. Moreover, we qualitatively motivate and empirically show that reducing recourse disparity across groups also improves fairness measures that rely on error rates. To the best of our knowledge, this is the first time that recourse capabilities across groups are considered to train fairer neural networks, and a relation between error rates based fairness and recourse based fairness is investigated.
Fused filament fabrication (FFF) has numerous process parameters that influence the mechanical st... more Fused filament fabrication (FFF) has numerous process parameters that influence the mechanical strength of parts. Hence, many optimization studies are performed using conventional tools and algorithms. Although studies have also been performed using advanced algorithms, limited research has been reported in which variants of the naked mole-rat algorithm (NMRA) are implemented for solving the optimization issues of manufacturing processes. This study was performed to scrutinize optimum parameters and their levels to attain maximum impact strength, flexural strength and tensile strength based on five different FFF process parameters. The algorithm yielded better results than other studies and successfully achieved a maximum response, which may be helpful to enhance the mechanical strength of FFF parts. The study opens a plethora of research prospects for implementing NMRA in manufacturing. Moreover, the findings may help identify critical parametric levels for the fabrication of custo...
This paper unveils a sustainable energy plan for optimal utilization of available electrical ener... more This paper unveils a sustainable energy plan for optimal utilization of available electrical energy resources for an energy-deficient village. The chosen village is Nangal, near Barnala, Punjab, India. Primarily, the requirements of electric energy are recorded and elaborated for around 450 households. Aiming this, the potential to harness electric power and its effective utilization has been identified from the available resources of energy: biomass, agriculture waste and solar photovoltaic (PV) technology. In order to achieve this, a hybrid renewable energy system (HRES) model is proposed whose performance is evaluated by implementing it in hybrid optimization model for electric renewable (HOMER) software. HOMER software provides optimal solution for a commercial biogas plant for catering cooking gas demand. Also, a coordinated solution for solar PV-operated water pumps used for irrigation, village water supply and solar PV street lights is presented and analyzed. In this way, the...
Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society
Fairness in machine learning is crucial when individuals are subject to automated decisions made ... more Fairness in machine learning is crucial when individuals are subject to automated decisions made by models in high-stake domains. Organizations that employ these models may also need to satisfy regulations that promote responsible and ethical A.I. While fairness metrics relying on comparing model error rates across subpopulations have been widely investigated for the detection and mitigation of bias, fairness in terms of the equalized ability to achieve recourse for different protected attribute groups has been relatively unexplored. We present a novel formulation for training neural networks that considers the distance of data points to the decision boundary such that the new objective: (1) reduces the average distance to the decision boundary between two groups for individuals subject to a negative outcome in each group, i.e. the network is more fair with respect to the ability to obtain recourse, and (2) increases the average distance of data points to the boundary to promote adversarial robustness. We demonstrate that training with this loss yields more fair and robust neural networks with similar accuracies to models trained without it. Moreover, we qualitatively motivate and empirically show that reducing recourse disparity across groups also improves fairness measures that rely on error rates. To the best of our knowledge, this is the first time that recourse capabilities across groups are considered to train fairer neural networks, and a relation between error rates based fairness and recourse based fairness is investigated.
Fused filament fabrication (FFF) has numerous process parameters that influence the mechanical st... more Fused filament fabrication (FFF) has numerous process parameters that influence the mechanical strength of parts. Hence, many optimization studies are performed using conventional tools and algorithms. Although studies have also been performed using advanced algorithms, limited research has been reported in which variants of the naked mole-rat algorithm (NMRA) are implemented for solving the optimization issues of manufacturing processes. This study was performed to scrutinize optimum parameters and their levels to attain maximum impact strength, flexural strength and tensile strength based on five different FFF process parameters. The algorithm yielded better results than other studies and successfully achieved a maximum response, which may be helpful to enhance the mechanical strength of FFF parts. The study opens a plethora of research prospects for implementing NMRA in manufacturing. Moreover, the findings may help identify critical parametric levels for the fabrication of custo...
This paper unveils a sustainable energy plan for optimal utilization of available electrical ener... more This paper unveils a sustainable energy plan for optimal utilization of available electrical energy resources for an energy-deficient village. The chosen village is Nangal, near Barnala, Punjab, India. Primarily, the requirements of electric energy are recorded and elaborated for around 450 households. Aiming this, the potential to harness electric power and its effective utilization has been identified from the available resources of energy: biomass, agriculture waste and solar photovoltaic (PV) technology. In order to achieve this, a hybrid renewable energy system (HRES) model is proposed whose performance is evaluated by implementing it in hybrid optimization model for electric renewable (HOMER) software. HOMER software provides optimal solution for a commercial biogas plant for catering cooking gas demand. Also, a coordinated solution for solar PV-operated water pumps used for irrigation, village water supply and solar PV street lights is presented and analyzed. In this way, the...
Uploads
Papers by Shubham sharma