tirthajyoti/deep-learning-with-python

Deep learning codes and projects using Python

Welcome 👋

Hello! This is Tirtha. I am an explorer.

• My LinkedIn profile
• My Medium articles on Data Science and Machine Learning
• My KDNuggets articles

Work

I am working as VP, AI/ML, at Rhombus Power Inc., where I am building exciting and critically important solutions with AI, Data, and Math.

Before this, I was a Data Science and Solutions Engineering Manager at Adapdix Corp, putting the power of AI/ML on the Edge for Industry 4.0 and next-generation Smart Factory.

Even before that, I was a Sr. Principal Engineer developing power semiconductor technologies and applying AI/ML for semiconductor product/tech deveklopment at ON Semiconductor, also known as onsemi.

At its core, I translate customer business problems into data-driven problems and help build solutions.

Currently...

• 🔭 Writing a book with Apress/Springer on the topic of "Productive and Efficient Data Science". You can pre-order the book here.

• 🔭 I’m currently working on: lectures/workshops, courses, and spreading knowledge on machine learning/statistical modeling. In particular, I serving as the Track Chair of "AI Optimization" track for the ValleyML AI Expo 2021. Also, I am developing course content for the ValleyML Fellowship program.

• 🌱 I’m currently learning: ML flow management tools, Ray serve and distributed computing, and how AI/ML applies to the various aspects of the Industrial IoT sector.

• 👯 I’m looking to collaborate on: Data science/ML books. Probably will use Jupyter Books and Leanpub platform

Books, lectures, articles

I publish highly-cited articles regularly on data science and machine learning topics, on leading platforms such Towards Data Science, KDNuggets, and Analytics Vidya.

I also teach IEEE/ACM workshops on data science/ machine learning.

My first data science related book Data wrangling with Python was published on February, 2019. In future, I wish to self-publish a second book about Hands-on mathematics/statistics for data scientists.

Skills

Open-source

My open-source projects span the topics of,

• general data analytics,
• machine learning,
• deep learning,
• computer vision and image processing,
• math and statistics,
• synthetic data generation, etc.

I have published multiple Python packages related to data analytics and statistical modeling. See this page for my projects

Contribution to the technical community

Currently, in the organizing committee of ValleyML AI Expo 2021.

I served on the Technical Content Committee for the Open Data Science Conference (ODSC) West, 2020.

In 2015, I was elevated to the grade of Senior Member of IEEE for my contributions towards power electronics. I have authored/co-authored more than 25 peer-reviewed Transaction and Conference papers, 2 monographs/book chapters, and 4 U.S. Patents. Here is my Google Scholar Page.

I also serve on the technical program committee as Track/Topic chair in numerous IEEE conferences. I am the co-chair of the Semiconductor Committee of Power Supply Manufacturers' Association (PSMA).

deep-learning-with-python's Issues

HELP: InvalidArgumentError: Graph execution error

Hi, I'm trying to implement your code from your:

Deep-learning-with-Python/Notebooks/Keras_flow_from_directory.ipynb

repo to a different dataset but run into the following error "InvalidArgumentError: Graph execution error" with a load of text afterwards; when trying to fit the model. I've tried a number of things but cant seem to fix it.

The only change I've made is the dataset, which is working and reading fine up to this point. And i've commented out the target_size in the train_generator.

Thanks, Tom

This is the top part of the error:

InvalidArgumentError Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_264/102295305.py in
1 n_epochs = 30
2
----> 3 history = model.fit(
4 train_generator,
5 steps_per_epoch=total_sample/batch_size,

~\anaconda3\envs\BigData\lib\site-packages\keras\utils\traceback_utils.py in error_handler(*args, **kwargs)
65 except Exception as e: # pylint: disable=broad-except
66 filtered_tb = _process_traceback_frames(e.traceback)
---> 67 raise e.with_traceback(filtered_tb) from None
68 finally:
69 del filtered_tb

~\anaconda3\envs\BigData\lib\site-packages\tensorflow\python\eager\execute.py in quick_execute(op_name, num_outputs, inputs, attrs, ctx, name)
52 try:
53 ctx.ensure_initialized()
---> 54 tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,
55 inputs, attrs, num_outputs)
56 except core._NotOkStatusException as e:

InvalidArgumentError: Graph execution error:

Detected at node 'categorical_crossentropy/softmax_cross_entropy_with_logits' defined at (most recent call last): ...

when try predict

thanks for your awesome code .but when I try predict I got two errors
the first is ("Error when checking input: expected conv2d_15_input to have shape (200, 200, 3) but got array with shape (200, 200, 4)
") and the second is that ("input' has DataType uint8 not in list of allowed values: float16, bfloat16, float32, float64
")

Issue while Predicting

Hello sir,
I am able to train the model successfully using the code provided in "Deep-learning with Python". I saved the keras model as well.I am trying to predict with a test image of my own. I used the same code as yours but I am running into the error:
ValueError: Error when checking input: expected conv2d_input to have 4 dimensions, but got array with shape (1, 200, 200)

Any idea how to fix this?