fixed notebook

smellslikeml · Jan 8, 2020 · 257c786 · 257c786
1 parent d923d1a
commit 257c786
Showing 1 changed file with 2 additions and 2 deletions.
diff --git a/Seizure_Prediction_using_CNNs.ipynb b/Seizure_Prediction_using_CNNs.ipynb
@@ -41,7  41,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "The original dataset is comprised of .mat files each containing 10 minutes of 15 channel EEG readings. For simplicity, we train on one patient's data and assume that each channel is equally well representative of the signatures we expect to learn in distinguishing whether the segment precedes a seizure event or not. It would consume too many resources to use the whole 10 minutes, so we transform these files into 1 second snippets and compute a spectrogram with consecutive Fourier transforms using the [scipy.signal](https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.spectrogram.html#scipy.signal.spectrogram) library. This lets us create an image of the change of a nonstationary signal’s frequency content over 1 second. "
     "The original dataset is comprised of .mat files each containing 10 minutes of 15 channel EEG readings. For simplicity, we train on one patient's data and assume that each channel is equally well representative of the signatures we expect to learn in distinguishing whether the segment precedes a seizure event or not. We ultimately want to deploy this model on a resource limited wearable so we sample 1 second snippets from a single electrode before computing the spectrogram using [scipy.signal](https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.spectrogram.html#scipy.signal.spectrogram). This lets us leverage powerful and efficient CNNs on the vision problem after windowing and fourier smoothing the noisy and nonstationary signal over 1 second. "
    ]
   },
   {
@@ -173,7  173,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "We can see that there is a discenable pattern between the two types of events. Now we'll move on to preparing all the data in this manner."
     "We can see that there is a discernible pattern between the two types of events. Now we'll move on to preparing all the data in this manner."
    ]
   },
   {