Maryam Miradi, PhD

𝐀𝐈, 𝐏𝐡𝐨𝐭𝐨𝐠𝐫𝐚𝐦𝐦𝐞𝐭𝐫𝐲, 𝐚𝐧𝐝 𝐃𝐫𝐨𝐧𝐞𝐬: 𝐑𝐞𝐯𝐨𝐥𝐮𝐭𝐢𝐨𝐧𝐢𝐳𝐢𝐧𝐠 𝐒𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐚𝐥 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬

I have done my PhD detecting and quantifying the severity of damage of Asphalt of all main and secondary roads. The added value was hundred of millions of Euros.

One challenge though: data collection which is super expensive and time consuming.

Aren is a civil infrastructure SaaS company has solved this challenge with building a solution incl. 3 steps:

➊ Aggregate raw data from drones, scanners, photos and sensors

➋ High-resolution 3D digital twin of assets is build capturing milimetric details

➌ AI/ML models automatically detects and quantifies the severity of damage.

ℝ𝕖𝕧𝕠𝕝𝕦𝕥𝕚𝕠𝕟𝕒𝕣𝕪 𝕊𝕠𝕝𝕦𝕥𝕚𝕠𝕟:

The combination of AI, photogrammetry, and drones makes it possible to assess structural conditions, track changes over time, and predict future rehabilitation needs.

ℙ𝕙𝕠𝕥𝕠𝕘𝕣𝕒𝕞𝕞𝕖𝕥𝕣𝕪:

It is a technique that extracts accurate measurements from photographs. By analyzing the geometry, scale, and perspective in images, AI algorithms reconstruct three-dimensional models of structures. This enables engineers to assess structural conditions with precision and detail.

ℚ𝕦𝕒𝕟𝕥𝕚𝕥𝕒𝕥𝕚𝕧𝕖 𝔸𝕤𝕤𝕖𝕤𝕤𝕞𝕖𝕟𝕥 𝕒𝕟𝕕 𝕋𝕣𝕒𝕔𝕜𝕚𝕟𝕘:

Through photogrammetry, AI can quantify defects such as cracks and efflorescence, providing proactive maintenance planning and prediction of future rehabilitation needs.

𝔻𝕒𝕥𝕒 ℂ𝕠𝕝𝕝𝕖𝕔𝕥𝕚𝕠𝕟 𝕀𝕞𝕡𝕣𝕠𝕧𝕖𝕞𝕖𝕟𝕥𝕤:

Accurate and uniform data collection is crucial for reliable photogrammetry results. Drones, such as Skydio enhance data acquisition by capturing high-resolution images, especially in inaccessible areas.

✍️ Do you know other applications of this kind?

See source in the comments 👇

#machinelearning #artificialintelligence #ai #deeplearning #datascience #analytics #dataanalytics #ml #python

𝐔𝐧𝐥𝐞𝐚𝐬𝐡𝐢𝐧𝐠 𝐭𝐡𝐞 𝐏𝐨𝐰𝐞𝐫 𝐨𝐟 𝐆𝐞𝐨𝐬𝐩𝐚𝐭𝐢𝐚𝐥 𝐃𝐚𝐭𝐚: 𝟐𝟎 𝐏𝐲𝐭𝐡𝐨𝐧 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬 𝐓𝐫𝐚𝐧𝐬𝐟𝐨𝐫𝐦𝐢𝐧𝐠 𝐋𝐨𝐜𝐚𝐭𝐢𝐨𝐧-𝐁𝐚𝐬𝐞𝐝 𝐒𝐞𝐫𝐯𝐢𝐜𝐞𝐬 & 𝐁𝐞𝐲𝐨𝐧𝐝

Geospatial data refers to information that is linked to a specific place on earth, such as geographic coordinates or addresses.

𝔾𝕖𝕠 𝕕𝕒𝕥𝕒 𝕚𝕤 𝕦𝕟𝕚𝕢𝕦𝕖 𝕚𝕟:

➊ Location Context
➋ Spatial Relationships
➌ Coordinate System Reference
➍ Visual Representation
➎ Complex Structures
➏ Data Integration

𝕀𝕥𝕤 𝕄𝕠𝕤𝕥 𝕀𝕞𝕡𝕠𝕣𝕥𝕒𝕟𝕥 𝔽𝕦𝕟𝕔𝕥𝕚𝕠𝕟𝕤:

➊ Geocoding: Converting address to coordinates
➋ Spatial Query: Searching for features like proximity to a point.
➌ Intersection: Finding the overlap between features.
➍ Buffering: Creating a polygon around a point, line or polygon.
➎ Union: Combining polygon features into a single one.
➏ Dissolve: Merging polygon based on a common attribute.
➐ Overlay: Creating a new layer by combining layers.
➑ Raster to Vector Conversion
➒ Distance Measurement: Calculating the feature distance.
➊⓿ Projection Transformation: Changing projection of a layer.


𝕍𝕒𝕝𝕦𝕒𝕓𝕝𝕖 𝕚𝕟 𝕤𝕖𝕧𝕖𝕣𝕒𝕝 𝕗𝕚𝕖𝕝𝕕𝕤:

⌘ Location-based services, like navigation apps.
⌘ Urban planning and land use analysis.
⌘ Environmental monitoring and resource management.
⌘ Crime analysis

ℂ𝕙𝕒𝕝𝕝𝕖𝕟𝕘𝕖𝕤 𝕠𝕗 𝔾𝕖𝕠𝕤𝕡𝕒𝕥𝕚𝕒𝕝 𝔻𝕒𝕥𝕒:

➊ Real-time Updates
➋ Accessibility
➌ Data Visualization

𝐈 𝐟𝐨𝐮𝐧𝐝 𝟐𝟎 𝐁𝐞𝐬𝐭 𝐏𝐲𝐭𝐡𝐨𝐧 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬 𝐟𝐨𝐫 𝐆𝐞𝐨𝐬𝐩𝐚𝐭𝐢𝐚𝐥 𝐃𝐚𝐭𝐚:

📚 Pydeck (⭐ 11K)

WebGL2 powered visualization framework

📚 Folium (⭐ 6.1K)

Interactive maps

📚 Geopy (⭐ 3.9K)

Geocoding & reverse geocoding

📚 Geopandas (⭐ 3.5K)

Geospatial data in a pandas DataFrame

📚 Shapely (⭐ 3.2K)

Geometric operations

📚 Rasterio (⭐ 1.9K)

Reading/writing raster datasets (satellite imagery)

📚 ArcGIS (⭐ 1.5K)

ArcGIS for Python

📚 PySAL (⭐ 1.1K)

Spatial analysis (spatial statistics & econometrics)

📚 Fiona (⭐ 1K)

Reading/writing geo data formats (shapefiles, GeoJSON, GPX)

📚 Pyproj (⭐ 840)

Projections & transformations of geospatial data

📚 NetworkX

Analyzing/modeling network data (spatial networks)

📚 Cartopy

Creating maps and plotting geospatial data

📚 Gdal

Working with various geospatial data formats/projections

📚 Gevent

Asynchronous I/O and network operations for large data sets

📚 RTree

Indexing/querying geospatial data

📚 Descartes

Plotting geospatial data in Matplotlib

📚 PyQGIS

Working with QGIS GIS software from Python

📚 OSMnx

Working with OpenStreetMap data (downloading, analyzing, visualizing)

📚 Geojson

Working with GeoJSON data format

📚 Geohash

Encoding/decoding geo data to ASCII string format.

✍️ Have I forgotten any techniques or libraries?

Source in the comments 👇

#machinelearning #artificialintelligence #ai #deeplearning #datascience #analytics #dataanalytics #ml #python #gis

𝐓𝐢𝐦𝐞 𝐒𝐞𝐫𝐢𝐞𝐬 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬 𝐃𝐞𝐦𝐲𝐬𝐭𝐢𝐟𝐢𝐞𝐝: 𝐓𝐨𝐩 𝟐𝟎 𝐏𝐲𝐭𝐡𝐨𝐧 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬 𝐩𝐥𝐮𝐬 𝐓𝐲𝐩𝐞𝐬, 𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬 𝐚𝐧𝐝 𝐅𝐮𝐧𝐜𝐭𝐢𝐨𝐧𝐬

Time-series data is a sequence of data points collected over time intervals, allowing us to track changes over milliseconds, days, or even years.

📊 𝔸𝕡𝕡𝕝𝕚𝕔𝕒𝕥𝕚𝕠𝕟𝕤:
- Monitoring software systems: virtual machines, containers, services, applications
- Monitoring physical systems: equipment, machinery, connected devices (IoT), our bodies
- Asset tracking applications: vehicles, trucks, physical containers, pallets
- Financial trading systems: classic securities, newer cryptocurrencies
- Eventing applications: tracking user/customer interaction data
- Business intelligence tools: tracking key metrics and the overall health of the business

📈 𝕋𝕚𝕞𝕖 𝕊𝕖𝕣𝕚𝕖𝕤 𝕋𝕪𝕡𝕖𝕤:
- 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞 Time Series: only one variable is varying over time.
- 𝐌𝐮𝐥𝐭𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞 Time Series: multiple variables are varying over time.

⌛️ 𝟙𝟙 𝕕𝕚𝕗𝕗𝕖𝕣𝕖𝕟𝕥 𝕔𝕝𝕒𝕤𝕤𝕚𝕔𝕒𝕝 𝕥𝕚𝕞𝕖 𝕤𝕖𝕣𝕚𝕖𝕤 𝕗𝕠𝕣𝕖𝕔𝕒𝕤𝕥𝕚𝕟𝕘 𝕞𝕖𝕥𝕙𝕠𝕕𝕤:
1. Autoregression (AR): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
2. Moving Average (MA): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
3. Autoregressive Moving Average (ARMA): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
4. Autoregressive Integrated Moving Average (ARIMA): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
5. Seasonal Autoregressive Integrated Moving-Average (SARIMA): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
6. Seasonal Autoregressive Integrated Moving-Average with Exogenous Regressors (SARIMAX): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
7. Simple Exponential Smoothing (SES): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
8. Holt Winter’s Exponential Smoothing (HWES): 𝐔𝐧𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
9. Vector Autoregression (VAR): 𝐌𝐮𝐥𝐭𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
10. Vector Autoregression Moving-Average (VARMA): 𝐌𝐮𝐥𝐭𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞
11. Vector Autoregression Moving-Average with Exogenous Regressors (VARMAX): 𝐌𝐮𝐥𝐭𝐢𝐯𝐚𝐫𝐢𝐚𝐭𝐞

𝐈 𝐟𝐨𝐮𝐧𝐝 𝐭𝐡𝐞 𝐟𝐨𝐥𝐥𝐨𝐰𝐢𝐧𝐠 𝟐𝟎 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬 for forecasting, anomaly detection, feature extraction, and machine learning 𝐨𝐧 𝐭𝐢𝐦𝐞-𝐬𝐞𝐫𝐢𝐞𝐬 𝐫𝐚𝐧𝐤𝐞𝐝 𝐛𝐚𝐬𝐞𝐝 𝐨𝐧 𝐆𝐢𝐭𝐡𝐮𝐛 𝐬𝐭𝐚𝐫𝐬 and project-quality score.

📚Prophet (🥇36 · ⭐ 15K )

📚Sktime (🥇35 · ⭐ 6K)

📚tsfresh (🥇32 · ⭐ 7K)

📚NeuralProphet (🥇32 · ⭐ 2.7K)

📚STUMPY (🥇32 · ⭐ 2.5K)

📚pmdarima (🥇32 · ⭐ 1.3K)

📚tslearn (🥈31 · ⭐ 2.3K)

📚Darts (🥈30 · ⭐ 5.3K)

📚GluonTS (🥈30 · ⭐ 3.3K)

📚Pytorch-forecasting (🥈29 · ⭐ 2.5K)

📚StatsForecast (🥈28 · ⭐ 2.2K)

📚Streamz (🥈28 · ⭐ 1.1K)

📚Uber/orbit (🥉26 · ⭐ 1.6K)

📚pyts (🥉26 · ⭐ 1.4K)

📚NeuralForecast (🥉24 · ⭐ 1.1K)

📚greykite (🥉21 · ⭐ 1.7K)

📚TSFEL (🥉21 · ⭐ 590 · 💤)

📚seglearn (🥉21 · ⭐ 540)

📚tick (🥉21 · ⭐ 410)

📚Auto_TS (🥉18 · ⭐ 520)


✍️ Do you know other Time-series libraries or functions?

Source in the comments 👇

#machinelearning #artificialintelligence #ai #deeplearning #datascience #analytics #dataanalytics #ml #python

𝐌𝐚𝐬𝐭𝐞𝐫𝐢𝐧𝐠 𝐔𝐧𝐬𝐮𝐩𝐞𝐫𝐯𝐢𝐬𝐞𝐝 𝐀𝐧𝐨𝐦𝐚𝐥𝐲 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧: 𝟐𝟎 𝐓𝐨𝐩 𝐀𝐥𝐠𝐨𝐫𝐢𝐭𝐡𝐦𝐬, 𝐄𝐯𝐚𝐥𝐮𝐚𝐭𝐢𝐨𝐧 𝐌𝐞𝐭𝐫𝐢𝐜𝐬 𝐚𝐧𝐝 𝐏𝐲𝐭𝐡𝐨𝐧 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬

𝔸𝕟𝕠𝕞𝕒𝕝𝕪 𝔻𝕖𝕥𝕖𝕔𝕥𝕚𝕠𝕟:

Unsupervised anomaly detection refers to identifying rare or abnormal instances in data without prior knowledge of what constitutes an anomaly, also called Novelty Detection.

𝔸𝕟𝕠𝕞𝕒𝕝𝕪 𝔻𝕖𝕥𝕖𝕔𝕥𝕚𝕠𝕟 𝔸𝕝𝕘𝕠𝕣𝕚𝕥𝕙𝕞𝕤:

Here are all 20 Anomaly Detection Algorithms I could find and their Python Libraries:

📚 𝐒𝐜𝐢𝐤𝐢𝐭-𝐥𝐞𝐚𝐫𝐧

O͜͡ Density-based spatial clustering of applications with noise (DBSCAN)
O͜͡ Isolation Forest
O͜͡ Local Outlier Factor (LOF)
O͜͡ One-Class Support Vector Machines (SVM)
O͜͡ Principal Component Analysis (PCA)
O͜͡ K-means
O͜͡ Gaussian Mixture Model (GMM)

📚 𝐊𝐞𝐫𝐚𝐬/𝐓𝐞𝐧𝐬𝐨𝐫𝐅𝐥𝐨𝐰

O͜͡ Autoencoder

📚 𝐇𝐦𝐦𝐥𝐞𝐚𝐫𝐧

O͜͡ Hidden Markov Models (HMM)

📚 𝐏𝐲𝐎𝐃

O͜͡ Local Correlation Integral (LCI)
O͜͡ Histogram-based Outlier Detection (HBOS)
O͜͡ Angle-based Outlier Detection (ABOD)
O͜͡ Clustering-Based Local Outlier Factor (CBLOF)
O͜͡ Minimum Covariance Determinant (MCD)
O͜͡ Stochastic Outlier Selection (SOS)
O͜͡ Spectral Clustering for Anomaly Detection (SpectralResidual)
O͜͡ Feature Bagging
O͜͡ Average KNN
O͜͡ Connectivity-based Outlier Factor (COF)
O͜͡ Variational Autoencoder (VAE)

𝐁𝐮𝐭 𝐡𝐨𝐰 𝐝𝐨 𝐰𝐞 𝐤𝐧𝐨𝐰 𝐰𝐡𝐢𝐜𝐡 𝐦𝐞𝐭𝐡𝐨𝐝 𝐢𝐬 𝐛𝐞𝐭𝐭𝐞𝐫? 𝐖𝐞 𝐝𝐨𝐧’𝐭 𝐡𝐚𝐯𝐞 𝐥𝐚𝐛𝐞𝐥𝐬 𝐢𝐧 𝐔𝐧𝐬𝐮𝐩𝐞𝐫𝐯𝐢𝐬𝐞𝐝 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠, 𝐍𝐨 𝐠𝐫𝐨𝐮𝐧𝐝 𝐭𝐫𝐮𝐭𝐡.

The answer lies in using evaluation metrics that can help us determine the quality of our algorithm.

𝔼𝕧𝕒𝕝𝕦𝕒𝕥𝕚𝕠𝕟 𝕄𝕖𝕥𝕙𝕠𝕕𝕤:

➊ Silhouette score:

A high Silhouette score (close to 1) indicates that data points within clusters are similar, and that the normal data points are well separated from the anomalous ones.

➋ Calinski-Harabasz index:

Calinski-Harabasz Index measures the between-cluster dispersion against within-cluster dispersion. A higher score signifies better-defined clusters.

➌ Davies-Bouldin index:

Davies-Bouldin Index measures the size of clusters against the average distance between clusters. A lower score signifies better-defined clusters.

➍ Kolmogorov-Smirnov statistic:

It measures the maximum difference between the cumulative distribution functions of the normal and anomalous data points.

➎ Precision at top-k:

The metric calculates the precision of the top-k anomalous data points using expert domain knowledge.

Don't leave your unsupervised anomaly detection to chance because there are no labels.

See the links to Papers and Tutorials 👇

✍️ Do you know any other evaluation methods for Unsupervised Anomaly Detection?

#machinelearning #artificialintelligence #ai #deeplearning #datascience #analytics #dataanalytics #ml #python

𝐔𝐧𝐥𝐨𝐜𝐤 𝐭𝐡𝐞 𝐏𝐨𝐰𝐞𝐫 𝐨𝐟 𝐏𝐫𝐞𝐭𝐫𝐚𝐢𝐧𝐞𝐝 𝐌𝐨𝐝𝐞𝐥𝐬: 𝟐𝟎 𝐓𝐨𝐩 𝐏𝐲𝐭𝐡𝐨𝐧 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬 𝐚𝐧𝐝 𝐚 𝐃𝐞𝐞𝐩 𝐃𝐢𝐯𝐞 𝐢𝐧𝐭𝐨 𝐓𝐫𝐚𝐧𝐬𝐟𝐞𝐫 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠

Transfer learning is a powerful technique that allows us to reuse pre-trained models to solve new problems efficiently.

𝕋𝕣𝕒𝕟𝕤𝕗𝕖𝕣 𝕃𝕖𝕒𝕣𝕟𝕚𝕟𝕘 𝕊𝕥𝕖𝕡𝕤:

➊ Select a pre-trained model.
➋ Remove the last layer(s) of the model.
➌ Add new layer(s) to the model.
➍ Train the model on the new dataset.
➎ Fine-tune the model (optional)

𝕋𝕣𝕒𝕟𝕤𝕗𝕖𝕣 𝕃𝕖𝕒𝕣𝕟𝕚𝕟𝕘 𝕋𝕪𝕡𝕖𝕤:

𝐈𝐧𝐝𝐮𝐜𝐭𝐢𝐯𝐞 𝐭𝐫𝐚𝐧𝐬𝐟𝐞𝐫 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠: Using a pre-trained model to solve a similar task to the original task.

🅞 Instance Transfer
🅞 Feature Representation Transfer
🅞 Parameter Transfer

𝐓𝐫𝐚𝐧𝐬𝐝𝐮𝐜𝐭𝐢𝐯𝐞 𝐭𝐫𝐚𝐧𝐬𝐟𝐞𝐫 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠: Using a pre-trained model to solve the same task as the original task, but with different data distributions.

🅞 Domain Adaptation
🅞 Semi-Supervised

𝐔𝐧𝐬𝐮𝐩𝐞𝐫𝐯𝐢𝐬𝐞𝐝 𝐭𝐫𝐚𝐧𝐬𝐟𝐞𝐫 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠: Using a pre-trained model to extract useful features from the data without any labeled data.

🅞 Self-Taught Learning
🅞 Deep Generative

𝕀𝕟𝕕𝕦𝕔𝕥𝕚𝕧𝕖 𝕋𝕣𝕒𝕟𝕤𝕗𝕖𝕣 𝕃𝕖𝕒𝕣𝕟𝕚𝕟𝕘 𝔸𝕝𝕘𝕠𝕣𝕚𝕥𝕙𝕞𝕤:

๏ Convolutional Neural Networks (CNNs):
- VGG16
- ResNet
- Inception

๏ Recurrent Neural Networks (RNNs):
- LSTM
- GRU

๏ Language Models:
- BERT
- GPT

๏ Autoencoder-based Models:
- Variational Autoencoder (VAE),
- Generative Adversarial Networks (GANs)

๏ Ensemble-based Models:
- Stacking
- Bagging

𝕋𝕣𝕒𝕟𝕤𝕕𝕦𝕔𝕥𝕚𝕧𝕖 𝕋𝕣𝕒𝕟𝕤𝕗𝕖𝕣 𝕃𝕖𝕒𝕣𝕟𝕚𝕟𝕘 𝔸𝕝𝕘𝕠𝕣𝕚𝕥𝕙𝕞𝕤:

๏ Domain Adaptation:
- Adversarial Discriminative Domain Adaptation (ADDA)
- Domain-Adversarial Neural Networks (DANN)

๏ Multi-task Learning:
- DeepFM
- Wide and Deep

๏ Self-taught Learning:
- Sparse Autoencoder
- Deep Belief Network (DBN)

๏ Cross-lingual Transfer Learning:
- Bilingual Distributed Representations (Bi-DRL)
- Language Model Pretraining (LMP)

๏ Few-shot Learning:
- Prototypical Networks
- Matching Networks

𝕌𝕟𝕤𝕦𝕡𝕖𝕣𝕧𝕚𝕤𝕖𝕕 𝕋𝕣𝕒𝕟𝕤𝕗𝕖𝕣 𝕃𝕖𝕒𝕣𝕟𝕚𝕟𝕘 𝔸𝕝𝕘𝕠𝕣𝕚𝕥𝕙𝕞𝕤:

๏ Language Models:
- BERT
- GPT

๏ Vision Models:
- ResNet
- VGG

๏ Audio Models:
- WaveNet
- MelNet

๏ Graph Models:
- Graph Convolutional Networks (GCNs)
- Graph Autoencoder

๏ Reinforcement Learning Models:
- Proximal Policy Optimization (PPO)
- Deep Q-Networks (DQNs)

𝐏𝐲𝐭𝐡𝐨𝐧 𝐋𝐢𝐛𝐫𝐚𝐫𝐢𝐞𝐬 𝐟𝐨𝐫 𝐓𝐫𝐚𝐧𝐬𝐟𝐞𝐫 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠:

📚TensorFlow
📚Keras
📚PyTorch
📚 Scikit-learn
📚 FastAI
📚 MXNet
📚 Caffe
📚 Theano
📚 Hugging Face Transformers
📚 OpenCV
📚 TFLearn
📚 Torchvision
📚 TensorFlow Hub
📚 TensorBoard
📚 TensorFlow Serving
📚 AllenNLP
📚 GluonCV
📚 PyCaret
📚 TensorFlow Probability
📚 PyTorch Lightning

See the links👇

✍️ Have I forgotten any libraries or algorithms?

#machinelearning #artificialintelligence #ai #deeplearning #datascience #analytics #dataanalytics #ml #python