Deep learning for smart sewer systems

Gudaparthi, Hemanth; Johnson, Reese; Challa, Harshitha; Niu, Nan

doi:10.1145/3377815.3381379

Cited by 8 publications

(7 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The desired MR is therefore E (t 0 − 1) ≤ E (t 0 ), reflecting the increased error caused by the simulated outlier. � Completeness: We have recently developed a novel MR to cope with the missing 'Flow' data [10]. Our source test case selects three consecutive time steps such as…”

Section: Stationaritymentioning

confidence: 99%

“…Although our previous work has investigated the MRs related to completeness [10] and stationarity [8], the framework presented in Figure 3 enables systematic and comprehensive MR derivations grounded in the problem domain's physical observables. While the data assumptions analyse specific situations affecting the input domain, the MRs complement the analysis with the desired properties of the output domain.…”

Section: Stationaritymentioning

confidence: 99%

“…This results in the follow‐up test case, which we use t ′ to denote. The desired MR is therefore E ( t ′ − 1) ≤ E ( t ′), reflecting the increased error caused by the simulated outlier. Completeness: We have recently developed a novel MR to cope with the missing ‘Flow’ data [10]. Our source test case selects three consecutive time steps such as E ( t − 1) ≥ E ( t ) ≥ E ( t + 1), showing an error‐reducing trend when no missing data is encountered.…”

Section: Framework For Assessing Dl's Fitness For Purposementioning

confidence: 99%

“…Our experience shows that a DL algorithm's accuracy in predicting the CSOs must be evaluated by considering what and how the data could go wrong in the problem context . For example, we characterised DL's robustness in the presence of missing data from the sensor network [10], and also tested how the stationarity of MSDGC's multivariate time‐series data affected DL's performance [8].…”

Section: Introductionmentioning

confidence: 99%

“…A public service organisation that we collaborate with is the Metropolitan Sewer District of Greater Cincinnati (MSDGC) [8][9][10][11]. Our joint work is aimed at exploiting DL methods to predict combined sewer overflows (CSOs), a problem considered by the U.S. Environmental Protection Agency (EPA) to be the largest unaddressed risk to human health from the water infrastructure [12].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Deep learning's fitness for purpose: A transformation problem frame's perspective

Gudaparthi

Niu

Yang

et al. 2023

CAAI Trans on Intel Tech

Self Cite

View full text Add to dashboard Cite

Combined sewer overflows represent significant risks to human health as untreated water is discharged to the environment. Municipalities, such as the Metropolitan Sewer District of Greater Cincinnati (MSDGC), recently began collecting large amounts of water-related data and considering the adoption of deep learning (DL) solutions like recurrent neural network (RNN) for predicting overflow events. Clearly, assessing the DL's fitness for the purpose requires a systematic understanding of the problem context. In this study, we propose a requirements engineering framework that uses the problem frames to identify and structure the stakeholder concerns, analyses the physical situations in which the highquality data assumptions may not hold, and derives the software testing criteria in the form of metamorphic relations that incorporate both input transformations and output comparisons. Applying our framework to MSDGC's overflow prediction problem enables a principled way to evaluate different RNN solutions in meeting the requirements. K E Y W O R D S deep learning, deep neural networks, software engineeringThis is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Section: Stationaritymentioning

confidence: 99%

Section: Stationaritymentioning

confidence: 99%