Embedded Files
Bereket Kibru Wolde
Highlights
Highlights
USGS Lidar Data Engineering
USGS Lidar Data Engineering
A python module that can be used to fetch, visualize, and transform publicly available satellite and LIDAR data and interface with USGS 3DEP and fetch data using their API.
Overview of the data source and formats
The USGS 3D Elevation Program (3DEP) has a collection of elevation data that is collected in the form of light detection and ranging (LiDAR) data over the conterminous United States, Hawaii, and the U.S. territories, with data acquired over an 8-year period. This dataset provides two realizations of the 3DEP point cloud data. The first resource is a public access organization provided in Entwine Point Tiles format, which is a lossless, full-density, streamable octree based on LAS zip (LAZ) encoding. The second resource is Requester Pays of the same data in LAZ (Compressed LAS) format. Resource names in both buckets correspond to the USGS project names. Have 1598 regions.
Features:
Features:
Download point cloud data from the EPT resource on AWS cloud storage.
Terrain visualization
Data transformation
Technologies used:
Technologies used:
Python, PDAL, Laspy, Geopandas, Shapely, Pydocs.
Brand Impact Evaluation
Brand Impact Evaluation
A/B testing is a user experience research methodology. A/B tests consist of a randomized experiment with two variants, A and B., which are identical except for one variation that might affect a user's behavior. It includes the application of statistical hypothesis testing or "two-sample hypothesis testing" as used in the field of statistic
Brand Impact Evaluation
Case Overview
SmartAd is a mobile-first advertiser agency. It is running an online ad for a client to increase brand awareness. The company provides an additional service called Brand Impact Optimizer (BIO), a lightweight questionnaire, served with every campaign to determine the impact of the ad they design.
Objectives
The task at hand is to design a reliable hypothesis testing algorithm for the BIO service and determine whether the recent advertising campaign resulted in a significant lift in brand awareness.
Important facts about the data
The exposed group seems to have slightly more counts of yes than the control group.
On the first day of the experiment, there was a higher user engagement.
A higher number of the participants who respond use chrome mobile
Methods
Classic A/B testing -Hypothesis
Sequential A/B testing - Hypothesis
A/B testing with Machine Learning
Conclusion
The classical test shows a 1.8% lift in brand awareness which is lower than the minimum detectable changeset and then we need more data to reach a conclusion using Classical A/B testing. The Machine Learning approach indicates the experiment (exposed and control user groups) feature doesn't have much significance. All this implies is that even if there was an increase in brand awareness, it is still not significant enough from the business perspective and we need more experiments before we reach a conclusion.
Swahili Speech To Text
Swahili Speech To Text
Build and serve end-to-end speech recognition systems for Swahili using various deep learning architectures that transcribe Swahili speech to text.
Approach:
Applied audio data preprocessing techniques like resampling, data augmentation by adding noise, changing the speed and pitch of the audio.
Cleaned the transcriptions by removing punctuations and non-alphabetic characters.
Extracted features from the audio data using log Mel spectrogram.
Built a model based on CNN and Bidirectional RNN that accepts Swahili audio data and transcribes it to text
Page updated
Google Sites
Report abuse