Even though, ML and DL community is growing the majority of growth is happening in the direction of Science (i.e., algorithms and code). Off late we are seeing lot of No code and Low code analytical platforms popularly known as AUTOML/DL tools which essentially automate the data exploration and modelling and deployment process.
But Data Science is not only about building the Models and optimization it is more of understanding the problem, data and its characteristics to understand the problem and hence provide the solution. Therefore, there is lot to do with the data( As Data Science is becoming more of data Driven). Hence we need a clear understanding of the data. The following are the some of the challenges which we face from data perspective while building ten Models.
- Data Quality
- Data Accuracy
- Data Completeness
- Data Consistency
- Data Timeliness
- Data Reproducibility
- Data Drift
- Scale
Therefore, Before Building the Model Please focus on these challenges and make sure that you have identified these and addressed these perfectly then only you can build the good model.
Source:https://pub.towardsai.net/common-challenges-in-machine-learning-and-how-to-tackle-them-cc29c47c5f24
